Supply Chain Management

lesson 1

12.01.2021

A supply chain consists of all parties involved, directly or indirectly, in fulfilling a customer request.
Let’s consider a real case: assume that you are buying a car (Fiat 500).
What is a supply chain? It is bundle of entities that interact with each other to satisfy your request. These entities can
be:
- Suppliers
- Manufacturers
- Retailers
- Logistic service providers
- The costumer
Within a car, a thousand of pieces are used.
On average, 90% of the value of the car is determined by the value of components which came from suppliers and
only 10% are manufactured by the car company (manufacturers).
Then, we have the car dealer, the retailer, which is an operator not involved in the production but specialized just in
the re-selling of the product.
Lastly, there are the logistic service providers, that are companies specialized in warehouses and on transportation
management. These activities, that are technical ones, are usually outsourced for two reasons:
1) In order to be efficient in warehouses and transportation management, it’s important to be able to deal with
great quantities. In order to do so, it’s fundamental to put together products of multiple manufacturers and
create therefore huge warehouse; the advantage can be obtained through the economies of scale, by
reducing the level of costs. The same idea can be applied to transportation management.
2) There are some technicalities, that are sometimes due to some legal constraints (like pharmaceutical
products’ warehouses cannot contain also other types of products, due to the law).

Þ How do they interact with each other? They interact exchanging information and materials with each other
in order get what they want (raw materials, semi-finished products, end products), with the quantity and
timing desired.

For example: supplier of packaging, manufacturer as producer of pasta (Barilla) and retail is a supermarket chain.
The process starts from the retailer that places an order to the manufacturer (red line, first flow of information), if the
manufacturer doesn’t have the product available for delivery, he will ask (second flow of information) to his supplier
what is needed for the production. From this point the flow, this time of physical materials and products, goes back,
from the supplier to the manufacturer and then to the retailer.
• Information flow goes from downstream to upstream
• Physical flow goes from upstream to downstream

1
An actual example of how the supply
chain works: Protecter&Gamble and
Walmart.
In this example, Walmart provides the
product, pricing and availability
information, to the customer. The
customer transfers funds to Walmart.
The latter conveys point-of-sales data
as well as replenishment orders to the
warehouse or distributor, that transfers
the replenishment order via trucks back
to the store. Walmart transfers funds to the distributor after the replenishment. The distributor also provides pricing
information and sends delivery schedules to Walmart, that may send back packaging material to be recycled. This
example shows that the customer is an integral part in the supply chain. In fact, the primary purpose of any supply
chain is to satisfy customer needs and, in the process, generate profit for itself.
The left side is what happens upstream, while the right one shows the downstream level.
Pactiv corporation: bottle of pharma as primary packaging.
Paper Manufacturer: as supplier for secondary packaging.
Timber Supplier: provides raw materials necessary.
Moreover, there are other suppliers such as Chemical Manufacturer and Plastic producer for packaging.
o Primary packaging: packaging in which the product is put (like plastic packaging).
o Secondary packaging: boxes used for transportation.

Which are the activities that should be carried out internally by a manufacturing company?
- Vendor rating: by comparing many potential suppliers, the company has to quantify vendors’ values and
then according to the rating the company selects the best one.
- Transportation of materials from vendor to producer
- Quality control of incoming goods: the quality evaluation of incoming goods can have different outcomes:
positive, so the control is passed, or negative, so the delivery is rejected.
- Materials handling in the warehouse: once the incoming goods are accepted, they are placed in the
warehouse.
- Capacity requirement planning: do we have enough capacity to satisfy the demand? To answer this question
1) forecast the future demand 2) can I produce that quantity? If I can’t, I can hire new workers or ask for
more time, or I can collaborate with subcontractors, that are external companies which will produce for you,
in order to fulfil that gap that I can’t produce.
- Managing subcontractors
- Production planning: once you have understood that you’ll produce 100 pieces and subcontractors 200
pieces, it’s important to construct a more detailed process, defining everyday activities involved in the
production process.
- Inventory management: once the final products are produced, they are put in the warehouse and the stock
level is set.
- Distribution planning: planning of the physical distribution, therefore the delivery of the product to the final
consumers.

Departments directly involved in SCM process:

1) Procurement: the procurement department that is in charge of selecting the supplier and managing the
relations with them.
2) Production: these department is in charge of managing the production process, therefore is the department
in charge of making the capacity requirement planning, so refers to understanding whether or not we can
produce. It is also in charge of selecting the subcontractors and managing them, planning the production, so
all the activities related with the production of the final goods.

2
 3) Logistics: downstream part of the supply chain, so to the management of the warehouse and the
transportation of the final products.
Sometimes, it can occur that there are internal conflicts among these departments.
If for example, in the producing of Barilla (pasta), the packaging is missing: the responsible for this shortage may be
the production manager or the procurement manager (1) it was a lot of time to get the packaging, but the manager
was not precise in respecting the plan, 2) or he selected a supplier company, but this company was not reliable).

Departments indirectly involved in SCM process:

1) Accounting and finance: Problems may arise with this first department, since it’s normal to anticipate money
(buying raw materials, financing activities etc.), through the so-called net working capital, which depends on
account receivable (money you expect from clients) + inventory – accounts payable (money you have to give
to your supplier), that the company obtains through the bank. How can you reduce your net working capital?
By changing one of the three components (accounts receivable, inventory and accounts payable):
- Accounts receivable: usually clients pay in 60 days, what can I do is to ask to pay the company earlier (shorter
time, however this is not easy, so usually this solution is skipped.
- Accounts payable: it’s possible to reduce it, by increasing the amount of time (I’ll pay supplier later), but it’s
not easy too, so this solution is usually skipped as well.
- Inventory: reduce the amount of inventory is the best solution. The negative implication, however, is that if
you receive an order it’s not easy to fully satisfy it because of the shortage of the stock. Therefore, it can be a
conflict between accounting & finance and logistics, that has the duty of fulfilling the order.
2) Marketing and sales: previously, we identified the so-called activity “capacity requirement planning” that
compares the demand and the available capacity. The marketing and sales department is responsible for
future demand forecasting, since it is closer to customers. However, it is not always accurate since it’s not
easy to estimate demand. Suppose that a company forecasts demand for 2021 to be at 1200 pieces, but the
actual one is instead 1500 pieces: what happens? The company will be out of stock and the responsibility falls
to the factory, since the demand estimation is not always accurate – this is something that is known.
Moreover, the factory should be flexible to quickly rearrange production to satisfy the actual level of
demand.
3) Design/R&D/New Product Development: departments in charge of product innovation. This department
may have an impact on SCM activities, since it always wants to innovate products, increase quality etc. The
production and procurement managers are not happy of this, because factories are requested to be efficient,
to change every time the features of products. High level of innovation brings high level of production costs,
causing conflicts.

Priorities to pursue when designing and managing a SC depend on the nature of the product According to Fisher,
there are 2 typologies of products:

• Functional Products, which are generally commodities; they are characterized by a stable demand pattern
and, therefore, their sales can be easily forecasted. For instance, for commodities (iron, nylon, paper, raw
materials etc.) competition is on price. They are characterized by stable demand pattern; therefore, sales
forecasting is quite easy.
Commodity is not the only type of functional product. Let’s consider toothpaste: even it is not a commodity
and each toothpaste have different characteristics, when one is discounted the consumer will usually opt for
that one, because in our mind each toothpaste works and performs its goal. Therefore, it behaves as a
commodity even if it is not.
• Innovative Products, with high innovation rates; they are subjected to a quick obsolescence due to fashion
trends or technological change. They are characterized by a broader products range and since they are
affected by fashion trends, they stay in the market for a very short time. For example, technologies
innovations (like phones) or Zara items. They do not compete on costs but on quality and innovation.

3
Examples:

Functional products... By reducing of 2 weeks the stock coverage of a Campbell soup, the retailer can enjoy
a saving equal to 1% of the selling price. The margin that it can earn from the soup is equal to 2% of the unit price.

Þ Demand is stable, product is always the same, it is a sort of commodity, competition is therefore on price.
!"#$!%# '#(!)' *"#$ +,# -#!$
Stock coverage: !"#$!%# .+*/0
Let’s suppose that for the Campbell soup, the demand over the year is 1200 and the average stock is 100.
1233
Stock coverage: = 120, therefore one month. Therefore, with the units in the stock I can satisfy the
133
demand of one month, so four weeks. If you can reduce your stock level for four weeks down to two weeks,
you can get a saving in terms of financial costs, that is 1% of the selling price. Is that a lot? Would you do
that? It depends, if you have a unit margin that is huge, like fashion results, 1% is not a high result. However,
we’re dealing with soups, therefore if the company saves 1%, it gets an incredible result.
In a situation like these, efficiency, cutting costs, is fundamental.

Innovative product... According to a survey conducted at Macy’s, a well-known department store, 25% customers
leave the store without having bought the item they wanted because it was out of stock.
Þ Fashion products, the unit margin is very high – which means, the shop buys a product from the supplier
which value is 50 and sells it to the customer at 100. By reducing the stock level, you save 1%, is that a good
improvement? No, even though you save money, you reduce the stock, therefore by increasing the possibility
of being out of stock. This causes a great loss in revenues – in order to save 1%, you may lose 100%. The
priority for fashion industries is having the product when it is requested by customers, which generates a lot
of revenues.

To meet these two different objectives, a specific SC should be planned.

For functional products, it’s fundamental to have an efficient supply
chain, instead for an innovative product, the reactive supply chain is the
best option.

According to Fisher, the two SC typologies have the following features:

• Reactive Supply Chain: It is able to promptly respond to market

needs, ensuring a quick availability of the product, thus minimizing stock-outs, even though this can involve
some production and distribution inefficiencies.
• Efficient Supply Chain: It is able to minimize the total logistic cost through an appropriate organization of
production and distribution activities.

In practical terms, these two chains are organized in the following way:

Efficient Supply Chain: to minimize costs through an efficient organization of production and distribution activities

o Production: High saturation of the plant – full saturation of the production lines, which means that the
production capacity equal to the quantity you have to produce, so that saturation is 100% and you do not
waste any resources.
o Low inventory level: the higher it is, the higher the financial costs.
o Selection of suppliers on the basis of cost and quality: therefore, given the quality you want to reach, you
select the cheapest supplier
o Lead time to be cut, but with no impact on costs: the lead time is the average time to produce an item. Being
very fast in producing, brings high level of benefits. However, the production system must be revised and
causes multiple costs. So if the level of costs is too high, it’s better to be slower in producing.

4
Reactive Supply Chain: Main goals is to be able to promptly respond to market needs, ensuring a quick availability of
the product, thus minimizing stock-outs, even though this can involve some production and distribution inefficiencies

o Production: Excess production capacity – low saturation. The product is volatile, which means that every
week the request changes. In order to satisfy the demand, a high level of production capacity is needed. The
implication is that there is a low saturation, so inefficiency.
o High inventory level: in order to satisfy the demand, I need a high inventory level.
o Selection of suppliers on the basis of speed and flexibility
o Lead time: to be cut, even though with unfavorable impact on costs – lead time reduction is a priority, to
meet demand it’s fundamental to produce very fast, even though a lot of money are requested.

ZARA CASE STUDY

1) Do you think Zara’ SC is properly organized?

- Zara produces innovative products, and it has a reactive supply chain.
Zara produces a wide range product, it is able to identify new trends, colors, fabrics and in a few days, it is
able to reach the market. This is conducted on a weekly basis. Quality is not excellent, but price is quite
cheap as well.
2) What are the pillars that make Zara so excellent?
a. Innovation process: it is able to detect the new trends, by analyzing very carefully fashion
magazines, fashion shows, so any place in which it is possible to get information pertaining to the
fashion market. Moreover, they need information to what is sold in the market, which can be
obtained by the vertical integration: they sell their own products, in their own shops. They collect
data on their sell-out, so on what is sold by them on the market by each shop.
Another source of information is the one of “shop assistants”. Shop assistants have been trained in
Zara to identify relevant information that has to be reported to design department. Let’s assume for
example that we work in a shop that doesn’t yellow items and everyday customers request yellow
clothes. If we rely only on the sell-out data, we’ll miss this information. Therefore, if the information
is owned by the shop assistants, it should be passed to the design department. Every day, at the end
of the working activities (daily basis), they meet to decide whether there is information that is
interesting and should be reported to the design department.
b. Speed and flexibility of production: Zara’ SC is completely different from others fashion industries’
supply chain. Usually, in fashion industries, production is outsourced to subcontractors, because
there are a lot of changes in the production of fashion items, therefore it is difficult for a brand to
have workers and machinery that are always able to perform new items, different between each
other, according to trends. Therefore, it’s better to buy production capacity from subcontractors.
However, the negative implication is that the time required for production is very slow – long lead
time. Zara model, on the other hand, requires speed and very short lead time; therefore, it has a
very unique supply chain.
Zara’s supply chain:

1) They do 50 – 60 percent of their manufacturing in advance versus 80 – 90 percent by competitors.

2) Zara can deliver garments to stores worldwide in just a few days: China – 48 hrs; Europe – 24 hrs; Japan – 72
hrs; United States – 48 hrs. It uses trucks to deliver to stores in Europe and uses air transportation to ship
clothes to other markets.
3) Zara buys large quantities of only a few types of fabric (just four or five types, but they can change from year
to year), and does the garment design and related cutting and dyeing in-house.
4) The company purchases raw fabric from suppliers in Italy, Spain, Portugal and Greece. And those suppliers
deliver within 5 days of orders being placed.
5) A relevant amount of textile is produced internally, by factories owned by Inditex.

5
 6) The 11 Zara owned factories are connected to the Cube by underground tunnels with high-speed monorails
(about 200 kilometers or 124 miles of rails) to move cut fabric to these factories for dyeing and assembly into
clothing items.
7) They employ about 3,000 workers in manufacturing operations in Spain at an average cost of 8.00 euros per
hour compared to average labor cost in Asia of about 0.40 euros per hour – however, they prefer speed to
costs.
8) Zara factories in Spain use flexible manufacturing systems for quick change over operations.
9) Opposite to its competitors, most of Zara’s production is in Europe: 50% of all items are manufactured in
Spain, 26% in the rest of Europe, 24% in Asia and Africa.
10) Compared to its competitors, Zara is highly vertically integrated, in order to minimize the time necessary to
move semi-finished products around.
11) Zara intentionally has some extra-capacity in its factories, so as to be in the possibility to seize market
opportunities by quickly producing what is required by the market.

lesson 2

15.01.2021

Performances of the Supply Chain: SC performances can be briefly divided into the following:

• Efficiency (total logistic costs): it means minimizing the total logistic-related costs;
• Effectiveness (service level): it means delivering to the client the product that he/she has asked for, with the
right quantity, in the right date, in the right place

Efficiency: we refer to the total logistic costs, that is the sum of four elements:

1) Transportation: you can use rail, sea, air etc. transportations. However, it’s important to make some trade-
offs between costs and speed: if you choose air transportation, costs will be high but at the same time
delivery is very fast. On the other hand, sea transportation is cheaper but slower.
2) Holding costs: costs that you incur by stocking a product, which is a financial cost that depends on the units’
value and the weighted average cost of capital of products. The higher the stock level, the higher the holding
costs. Example: let’s assume that we are a firm
3) Warehouse costs: it’s the cost of the facilities (energy, water etc.) and the costs related with it. There can be
cases the warehouses don’t belong to the firm: the firm wants to use just 200 squared meters of another
company’s warehouse. The logistic section of this firm, will decide the costs, including warehouse costs.
4) Administrative costs: the costs associated to the cost of the information system that you need to manage in
supply chain activities and the cost of people working in the firm (example: salaries of the accounting
department).

Effectiveness: we refer to the service level, that is a bundle of performances, the most relevant ones are:

1) Speed of the delivery. In many companies when it’s fundamental, like in Zara, where is it used as the primary
competitive advantage to succeed in the marketplace.
2) Dependability: ability to deliver on time, therefore, to meet the due date. Let’s consider a radar, used in
airports, and suppose that Malpensa needs a new one. Once the order is placed, the time of the delivery may
take even two years, since it’s a product that has to be produced according to the features needed in that
specific place. Once it is produced, some trials are made etc. Putting all these activities together, Malpensa
will wait at least two years. It’s not possible to make it faster. Therefore, when a company places an order
like this, it knows that it takes a lot of time, but the time has to be respected. Being on time is extremely

6
 relevant. An earlier delivery, that in some industries may occur, it may be an issue because there may not be
placed to stock the product.
3) Availability and Completeness:
a. Availability: it means having a product when the customer orders it. If the company has it in its
catalogue but it is not available at the time of the order, this will make the customer very
disappointed. Therefore, availability is extremely important in some industries, especially for
commodities.
b. Completeness: Let’s assume that you make an order to Amazon: a mobile phone and the textbook.
You receive the order, but on two different days: this may not be a problem, you do not need them
together to use them. Instead, if I make an order on Esselunga, by selecting all the ingredients
needed to make a cake: sugar however doesn’t arrive, since the product wasn’t present in the stock.
Therefore, a problem in availability causes problems in completeness. There are cases in which
completeness is fundamental, that is the ability to deliver all the items that have been ordered on
time.
4) Flexibility: The possibility to change an order once it has been placed. For instance, on Amazon you can
change the order because there is a number of hours between the moment in which you place an order and
the moment in which they fit the product in the warehouse. During this time, you can change the products or
even eliminate it. Let’s consider a fashion industry, where a brand has to decide how many items has to
produce for the next season: this is really difficult, because it’s hard to forecast sales. Therefore, in some
cases, especially when the forecast is not easy to make, flexibility is very important because the
customer/company appreciates the possibility to modify the orders.

Total logistics costs: some data

• In North America, Europe and Asia Pacific, logistics expenditure accounts for 11% of the GDP. In Latin
America it is 14%. Until 80’s, this value was nearly 20% even in more advanced Countries. In Italy it’s nearly
11%. The lowest values (lower than 10%) are reported by highly developed Countries (e.g. Germany, Japan,
France, UK, U.S.), where the awareness about the relevance of logistics is very high since long. Countries as
India and China report much higher costs (17% and 21% of the GDP).

Þ We can say that logistics accounts for the larger part of GDP and that the variability of these values depends
mostly on the level of infrastructures (if they are underdeveloped, logistic costs are higher and vice versa).
Moreover, the percentage depends on managerial practices level that are used in logistics; therefore, in
countries like Italy, UK, USA (developed countries), best practices in logistics are adopted. In fact, thanks to
the level of efficiency it is possible to achieve, logistic costs are lower.

Total logistic costs as a % of a company’s turnover

Let’s make a comparison, starting with cement. Why logistics in this industry is so expensive? Because of the
transportation and warehouse costs. Transportation is very expensive because the products are heavy: the main

7
drivers therefore are size and weight. The more a product is heavy, the more it is expensive to transport it. Moreover,
at the same time, you’re moving a cheap item, so this result that revenues are low compared to the size and weight
and compared therefore to the transportation costs. The implication specific to this industry (cement), is that when
you manage the transportation of cement, you try not to make huge distances; if the distance is really high, however,
the costs of transportation would be so high to not generate profit. This is the reason why cement industries are very
widely spread on the territory.

On the other hand, computer supply’ logistic costs are very low. Let’s consider an iPhone, which is very small and
weightless. This causes a low cost of transportation costs, but at the same time a high level of revenues. Therefore,
the transportation costs are low, but there is a high level of revenues.

Total logistic costs: breakdown

This graph makes a comparison of logistic costs between EU and US. There is a relevant difference in transportation
costs: in US accounts for 49% of logistic costs, while in EU accounts just for 40%. This is due, of course, to the longer
distances of the US territory.

To compute total logistic costs, we need the activity-based costing, but since this is not an accounting course we’re
not going to focus on computation.

Service level includes speed, dependability, completeness and availability and flexibility. Let’s see how to compute
them:

Service level: speed

Example: In this figure, you can see an arrow with a timeline of an order that was placed on the 1/06 (Order Date –
OD) and on the 10/06 the supplier has to make the delivery (Due Date – DD). However, sometimes deliveries are late,
so the first actual delivery date is 15/06 (ActuaL Delivery Date – ADD1) and the second one is 20/06 (ADD2); this
occurs because sometimes the order arrives in two different dates. For example, I order at the same time on Amazon

8
a book and a pen, but they arrive on two different days, therefore I’ll have ADD1 and ADD2. To calculate the speed, I
use the two formula reported in the picture. If there are two actual delivery dates, consider always the first formula,
that is a sort of average. Let’s suppose a customer is looking for a supplier that meets a delivery of a product in x days,
then in this case the second indicator is used. For example, suppose that a client wants to receive the order no later
than 18 days: in order to understand whether we’re able to do it, we look at previous orders and we see that just two
orders out of three were performed within 18 days, the third one was fulfilled on the 20th day. Therefore, the second
indicators will be (2/3) x 100=66%.

Service level: dependability

Dependability means being able to deliver on time. It is computed using the formula and if we have two actual
delivery date, we just consider the first actual delivery date and we use the first formula, which is a sort of average.
How to compute the first formula? Let’s suppose I have three orders: the first two were on time, the third one had a
delay of 15 days. Therefore: (0+0+15)/3 = 5 days in average of delay. How to compute the second formula? Number of
orders with a delay, in our example, would be 1, and the total number of orders would be 3. Therefore, 33%.

Example: Order 1 is computed with a delay of five days, while order 2 in advance of 5 days. The advance, from a
numerical point, is -5 (ADD-DD). Dependability: (+5-5):2=0, the average delay in delivery is 0. However, this is not
correct, the company does make delays. To solve this problem:

From a numerical viewpoint, since the advance is a problem, you will consider the advance in absolute term, which
brings -5 to +5. The average delay in delivery turns (5+5)/2 = 5, therefore on average there is a time lag of five days.

9
Now consider a case in which a delay is not a problem, for example when you place an order to a supplier, and it
comes before planned: this is not a problem, but an opportunity, then the advance is considered as 0, you’re just
interested in actual delays. The average delay in delivery turns (5+0)/2=2,5 days.

Service level: completeness

Completeness is the ability to deliver jointly all the pieces requested, which is very important in different cases. Let’s
consider this table, where there are two orders. In Order 1, two products were requested, Product A and Product B.
For product A, 100 pieces were requested, and they were fully requested in the 1st delivery. For product B, instead, 50
pieces were requested, and they were delivered just on the 2nd order. Therefore, this order requires two delivery. For
Order 2, instead, both products (C and D) were fully delivered and jointly in the first delivery.

How to compute completeness? Completeness (1) = ½ x 100 = 50% (just the second order was fully completed in the
first delivery).

This indicator, instead, is more precise, since it is based on the lines. The lines correspond to the product/item
(product A, B, C and D, so four lines). In this case, product A, C and D were fulfilled in the first delivery out of four,
therefore 75%.

This third indicator is based on the number of pieces in the first delivery over the number of pieces requested.

Let’s suppose the supplier is a supplier of stationary products (tuner, pens etc.), so everything is needed in an office
and suppose that product A is a cheap product, like pencils (1$ per piece), while product B is very expensive, like toner
(50$ per piece). If we get paid based on the date of the delivery, that would be a product, since the most expensive
one is on delay. In order to quantify this issue, we need one more datum, the unit price, in order to understand the
value of the first delivery out of the whole order (e.g., for product A is 100$ etc.) and we divide the sum by the value
of all pieces requested. This is the following formula:

10
This means that we are delivering cheap items first and expensive items later; this has a bad impact on the economics
of our company.

This is the last indicator, that is based on number of deliveries to fulfill an order. For the first order we needed two
deliveries and for the second order we needed two as well. Therefore, by making an average the result is two.

Þ Because completeness often depends on the physical availability of stock in the warehouse, the analysis of
this performance is complemented with the computation of other indicators, which express stock
availability.
How do we measure stock availability? It is based on the concept of stock out, that occurs when we do not
have enough stock to satisfy the demand.

Service level: availability

Availability can be measured in terms of stockouts.

ü Product A: stock-out of 10 pieces

ü Product C: stock-out of 20 pieces
ü Product D: stock-out of 10 pieces

Consider that we’re controlling stock-out for product A for four weeks.

lesson 3

11
OTIF: On-time in Full indicator

Range of tolerance

ü Order 1: The delivery was on time, but it was not fully completed
ü Order 2: The delivery was completed, but it was not on time
ü Order 3: The delivery was on time, but it was not fully completed

Range of tolerance: the customer is still happy in a range of 4 days before the DD and 4 days after the DD. Therefore,
Order 2 can still be considered on time, due to the range of tolerance.

This means that completeness and punctuality sometimes have to be combined; through the OTIF it is possible to do it
and measure these parameters together.

Service level: flexibility

• Flexibility can be defined as the ability of a system to change itself, but under two conditions:
o speed of change
o cost of change
• A system is flexible if it can change quickly and with a limited cost

Example: in Amazon it is possible to change or cancel an order before the picking day of the order, without any costs.
This is flexibility. In other cases, it may be possible that some companies put a day limit until when it is possible to
modify the order: if the change is desired after that time limit, the company will not respond of the change – it is
inflexible. Flexibility can be measured following these two indicators:

12
The most flexible supplier is the one that has the lower number of days of advance to accept a change in the order
(first indicator). If the amount requested to be changed (second indicator), if it is too high, the supplier may not be
able to make the deliver or if it is too low, it would be too costly.

lesson 3

19.01.2021

CASE STUDY: CLOTHES AND MORE

1. Do you think that the two indicators provide a misleading view of the actual logistic performance of
the company?
§ The two indicators do not distinguish between either typology of client (fashion houses vs
retailers) or typology of orders (planned vs replenishment) thus, e.g., we cannot properly
assess speed. It makes sense to make a distinction between types of clients; in the
relationship between fashion houses and retailers, fashion houses have a high bargaining
power, since they are companies itself compared to retailers, that are shops. On top of this,
fashion houses are present both in upstream and downstream in the supply chain.
Distinction between planned and replenishment orders is fundamental, otherwise speed
cannot be assessed properly: planned orders require more time to be produced, compared
to replenishment orders that have to be completed in a shorter term.
§ Concerning the dependability performance, delayed planned orders that are delivered
before September 1st should not be considered critical (e.g., Order 1), since in the eyes of
the clients September 1st is the important date.
§ With the current approach to the computation of this indicator, incomplete orders can be
considered “on time” (e.g., Order 3), thus no relevance is given to completeness. On August
15, the majority of the pieces weren’t delivered to clients, they were delivered just on the
20th of September. If indicators of completeness were computed, this situation would had
emerged.
2. Would you compute some other indicators and, if so, which ones would you compute?
§ Yes!
§ In the matrix, a proposal is made:
the white box is unrealistic,
because fashion houses do not
make replenishment orders.
Replenishment orders are
important just for retailers.
§ What does the client desire when
he makes an order? (fashion
houses) The client is interested in
receiving complete delivery and on time. Speed is not important because replenishment
orders do not exist. Therefore, the best indicator is OTIF for fashion houses.
§ What does the retail desire when he places an order? (retailer)The same thing: the order to
be made on time and in a complete way. OTIF therefore, is still fundamental, but due to the
existence of replenishment orders, speed is important as well.

13
Planned orders’ computation:

In this table, 12 orders are planned, since they are delivered within the 1st of September and just three orders are
placed by fashion houses (green ones). For two of them, OTIF is computed, OTIF 1 and OTIF 2 (OTIF 3 is not computed,
because the range of tolerance about the delivery is not provided).

OTIF 1: Order 1’s due date was on the 15th of July,

but it was actually computed on the 20th; however, it
is still considered on time compared to September
1st. Order 5 is not on time, while Order 12 is on time
but not in full. In conclusion, only Order 1 is both on
time and fully completed. OTIF 2 considers
completeness: Order 1 100% completeness, Order 5
0% completeness, Order 12 66% completeness,
therefore the average is 55.3%.

Replenishments orders: Speed and Completeness

Speed is really fine: average 5 days and all the orders

are delivered in that amount of time (100%).

Completeness in terms of complete orders out of the

total is 66%, in terms of pieces delivered in the first
delivery is 71% and the stock-out rate (availability) is
25%. Stock-out rate can be computed in terms of
items: with these 3 orders, 4 different items are
been requested through replenishment orders. Item
C was delivered later, therefore it was not available
in the first deliver. Consequently, the stock-out rate
is 1 out of 4 (1/4), which is 25%. It’s important to
compute stock-out rates, because stock-out
availability is present when the production has been made with advance. It doesn’t make sense to compute it with a
“make to order” approach, as in the case of planned orders, because by definition the company doesn’t have any
stock. If you cannot produce all the volume that you planned is because you had some production constraints and you
have to postpone the delivery, therefore you’re not complete. Instead, if you have some problems with replenishment
orders, then in that case, is a stock-out problem, that usually occurs when forecast was wrong.

You can also add some other indicators like OTIF and mean delay in delivery, but the main ones are definitely speed
and completeness.

When you assess the logistic service, not all the indicators have to be computed, but focus on the specific case and
imagine on what the customers need – use the perspective of the clients.

14
 lesson 4

22.01.2021

DEMAND FORECASTING AND PRODUCTION PLANNING PART I

Demand Forecasting and Production Planning aim at supporting the decision-making process that refers to:
1) how much to produce: in order to decide how much to produce in a year, demand forecasting has to be
conducted, in order to understand first of all how much we want to sell.
2) when to produce: suppose that we are dealing with seasonal demand, I’ll
draw a chart that represents the seasonality: on the x axis there is time,
on the y axis there are pieces that will be sold. The shape of the demand
would be like that: high in summer, low in the rest of the year (blue curve)
and the total demand will be 1200 pieces in a year. Therefore, the total
production would be 1200. When to produce? Two alternatives may arise:
1) I’ll produce when the client requests products (red curve), 2) I’ll
produce an equal number of products over months (flat line), no matter
what the demand is. Of course, with this method, the company accumulates stocks that will be useful when
there is a pick of demand.
3) which production resources to use: suppose that you have to produce 1200 pieces, but you can produce just
1000 pieces. What can you do?
o Request overtime
o If the gap is rather large, you can think of some shifts for some months of the year
o Use of subcontractors
Therefore, the issue related to resources, is associated to the fact that sometimes companies do not have
enough large production capacity and have to understand how to fill the gap.

All these decisions are related to the so-called

manufacturing planning process, a complex process where
for each step there is a different production plan. What are
the differences among all these production plans?
1) Time horizon: at the top of each process, you have
long term planning, then medium and lastly term
planning.
2) Object of each plan: like in the Nutella case, each
step has its own object, and it becomes different
with the moving in timing (long-term, medium-term
and short-term).
Capacity planning: every time a company draws up a new production plan, before launching it, the company has to be
sure that it possesses the right amount of capacity to perform it.

Example: in the Nutella case, you can see the three different terms.
Long-term planning relates to the general decision pertaining to the
creation of a product (not how many pieces, just in tons), then in the
medium-term you decide with precision how many pieces and also
the timing and in the short-term the focus is on raw materials and
components.

15
DEMAND PLAN is a document that reports sales forecast for the future and it is drawn up by the department of a
company that has the highest knowledge of the market, which is usually the marketing and sales department. This
document reports two type of information:

1) The quantity that is likely to be sold over a certain period of time

2) The pattern of demand in order to understand:
a. Demand is with or without seasonality
b. Demand is with or without trend

This information is referred to the whole production volume or to product families – e.g., total amount of Nutella
(whole production volume); for an automobile company product family can be the models, with no regard to
horsepower, colours, etc.
Forecast has to be developed over the long run (generally, 1 year, divided into months or quarters), however it
depends on the industry, like in fashion industries it may be six months due to seasonality and trends.

Demand plan: an example

Demand plan of a company that produces two

products, A and B, which require different resources
and time. For example, suppose the company is a
fashion house that produces t-shirts and coats. Total
demand for product A is usually 7681 and for
product B 8449, and the expected demand is split
into months. The pattern seems to follow a
seasonality, since it increases during wintertime and
it quite low in the middle of the year.

The next step is capacity planning: Operations managers need to know future sales in order to check whether there is
large enough production capacity to meet market demand
This check has to be performed in 2 different steps:
1) the first check aims at stating whether, over the specified period of time (e.g., a year), the production system
has large enough capacity to manufacture all the units that the company is likely to sell.
2) the second check aims at comparing the demand peak with the capacity of the plant in every single month
(or whatever the period of time has been divided into).
In the previous example, the peak is in December, therefore the second check is conducted in that month.

Once we’ve carried out these two checks, we can move to the next step of the production planning process, that is
PRODUCTION PLAN.
The Production Plan:
• has to be drawn up by the Operations Department
• refers to the long run, and covers a period of time as long as the one considered in the Demand Plan (for
instance, a year divided into months)
• is referred to whole production volumes or Product Families, as the Demand Plan does
• states the quantities that have to be manufactured for every product family in each period (for instance, in
each month that the year has been divided into)

The difference between the demand plan and the production plan is the aim: the former aims at understanding how
much we will sell, the latter, how much we will produce in the next year, divided in the months.

16
If we do not have any constraint in the production capacity, we can choose between two alternative Production Plan:
Þ Level Production Plan, that keeps the production constant from period to period
Þ Chase Demand Production Plan, that adjusts production to reflect the fluctuations in demand
In order to choose among these two plans, the two checks previously described have to be passed.
Even though, if we do have capacity constraints it is sometimes necessary to draw up a Level Production Plan: when
the second check doesn’t pass, the chase demand production plan cannot be used, only the level one can be
implemented.

Let’s see the case in which it is possible to choose, let’s consider the two plans:

Level Production plan

Consider a case in which the production capacity is a little

bit lower than the peak of the demand. Then, by
implementing the level production plan, during the
months in which we will have a demand lower than the
production volume we will accumulate stocks that will be
used in October, November and December, during the
peak of the demand.
Advantages:
- Stable employment pattern: i.e. every single
month you have the same quantities of workers and don’t have to assume season worker. This is an
advantage because you don’t need to introduce people just for a short period of time which can be costly
and complex.
- High process utilization: if you have a capacity constraint, by using this plan, you are likely to use 100% of
your capacity every month. This is very important especially if you have a capital-intensive production like
Nutella, it is good to exploit the available production capacity as well as possible. Let’s assume that you are
producing Nutella with the production process explained before and let’s assume that you have to stop the
process for some hours and then restart again: this situation comports a great loss of money because during
the period of time in which the facilities don’t work you are not producing any item, you are just wasting
energy. This happens in the most capital-intensive operations in which is important to have a stable
production pattern, where the facilities work 24 hours a day, and if you stop the production there isn’t a high
process of utilization.
- High productivity à low unit cost: it is important to have a stable production pattern because with it you can
increase the level of efficiency.
- Minimized stock out costs: the cost you incur when you are out of stock. The stock-out situation arises when
you have some demand, and you are not able to satisfy it because you don’t have available product. With the
level strategy you have always some stock availability (brown area), that you produce in some periods of the
year, in order to increase the capacity available, so that when you reach the period of peak of demand, in the
warehouse there a lot of pieces to fulfill clients demand,
Disadvantages:
- Not feasible with perishable goods, or with products where fashion changes rapidly because you cannot keep
the product in stock form months, because after a short period the level of your product will be lower. This
happens also for technologic and electronic products, because the commercial level will be lower during the
time.
- High holding costs, which are the costs sustained when you have your warehouse full of products. It is a type
of financial costs, and if it they are very high, it would be a disadvantage to keep products in stock for a very
long time (very expensive).

17
Chase demand production plan
In the chase demand production plan, the production
capacity is a lot, therefore it is easier to satisfy the
peak of demand and every month it is possible to meet
the demand of customers.
Production we produce is the brown lines, which meet
the forecasted sales.

Advantages:
- Low holding costs: if I produce today, I’m going to sell tomorrow – therefore, there is no stock in the
warehouse.
- No risk of changes in fashion or obsolescence: since the company doesn’t hold any pieces in the warehouse.
Disadvantages:
- Higher risk of stock-out: since the company doesn’t have any stock, there is a great risk if the forecasted sales
are underestimated. Therefore, the risk of stock-out is higher in the chase strategy than in the level strategy.
- Low utilization and productivity: the company doesn’t fully utilize the total production capacity; therefore,
the company uses the facilities the less it can. It could produce more, use time to do that, since you can see
also through the graph the huge difference between the actual production capacity (black line) and the
production level (brow line).
- Not suitable for capital-intensive operations: more suitable for labor intensive operations, since with this
method you can increase your production capacity using overtime and paying workers just for that period of
time and then dismiss extra workers. Therefore, there is a sort of flexibility into production costs. On the
other hand, with capital-intensive production, it’s not possible to go overtime: you have to pay for all the
equipment and full exploit it.

Example: Level vs Chase Demand Production Plan

Let’s consider the analysis has been conducted and all the two models could be applied: in the below tables, it’s
possible to see the production for each month. For the level production plan, production would be a little bit higher
than the actual one, while for the chase demand it would be in line with the true one.

EXERCISES

Question 1
The demand of a product, affected by seasonality, is 200 units/month from January to June and the double from July
to December for a total of 3600 units. Knowing that:
• the yearly demand must be produced;
• each month is made of 20 standard working days;

18
 • the plant dedicated to the product can reach a max production capacity of 20 units per day;
• the inventory holding cost is 2 €/unit per month;
• the cost of running the plant (whether it works or not) is 1 €/producible output;
• the set-up cost to change the production volume is 2,000 €/set-up;

To minimize the yearly cost please:

Þ Select between a chase or a level plan

Þ Report the calculation to take the above decision

Solution

Þ Level plan assuming 300 units/month:

o Holding cost = 100 units/month x 6 months x 2 €/unit-
month = 1,200
o Differential running cost due to unsaturation=100
units/month x 12 months x 1€/units = 1,200€
o Total cost level plan= 2,400 €/year

Þ Chase plan, assuming to follow the demand trend:

o Differential running cost due to unsaturation = 200
units/month x 6 monthsx1€/unit = 1,200 €
o Set-up-cost = 2,000 €/set-up
o Total cost chase plan = 3,200 €/year

Question 2

Alfa is a manufacturing company that produces two items, A and B. For the next year, the sales forecasts of A and B
are reported in the Table below. They refer to the number of pieces of A and B.

• Assuming that the production of 1 piece of A requires 30 minutes and the production 1 piece of B requires
120 minutes, what is the Production Capacity that would enable the adoption of a Chase Production Plan?
• Assuming that Alfa has 240 working days in a year, 8 working hours each, compute the Available Production
Capacity and the Necessary Production Capacity in terms of pieces.
• If the operating conditions of this plant do not allow a large enough Available Production Capacity to
implement a Chase Production Plan, what is the Production Capacity that would at least enable the adoption
of a Level Production Plan?

Answer to Question 1
Check at the year level:
Necessary Production Capacity (year): (275pcs * 0,5h/pc) + (950pcs * 2h/pc) = 2037,5h

19
Check at the peak level:

Necessary Production Capacity (month): 222,5hours

Answer to Question 2
Unit of measurement: A
Conversion rate (B in terms of A): 2 hours / 0,5 hours = 4
Necessary Production Capacity (year) = 275 A + 950 B = 275 A + 950 B * 4 = 4075 A
Available Production Capacity (year) = (240d * 8h/d) / 0,5 h/A = 3840A

Answer to Question 3
Necessary Production Capacity for Level (pieces per year) = 4075 A
Necessary Production Capacity for Level (pieces per month) = 4075 A / 12 = 340 A
Necessary Production Capacity for Level (hours per year) = (275pcs * 0,5h/pc) + (950pcs * 2h/pc) = 2037,5h
Necessary Production Capacity for Level (hours per month) = 2037,5h / 12 = 170h

lesson 5

29.01.2021

DEMAND FORECASTING AND PRODUCTION PLANNING PART II

The Manufacturing Planning Process

Once the Production Plan has been drawn up, it is necessary to get further details about:

• how the quantity to be manufactured for each product family has to be split into specific items;
• how much has to be manufactured in shorter periods of
time.

This information is provided by the MASTER PRODUCTION

SCHEDULE

The MPS:

• has to be drawn up by the Operations Department

• it covers a period of time shorter than the Production Plan (generally from 1 to 6 months)
• is referred to an object more specific than the product family

20
A key issue in drawing up the MPS is the choice of the items for which the MPS has to plan the manufacturing
activities. This choice depends on the production planning approach of the company.

Structure of a lead time

To understand the different types of

production planning approach, it’s
fundamental to consider the lead time of a
product, total time necessary to produce an
item. The lead time is divided in four main stages:

1) Engineering activity: designing the product;

2) Purchasing activity: procure the main resources and materials from suppliers;
3) Production of components and sub-assemblies;
4) Final assembly: put together all the subcomponents.

According to the industry in which the company is engaged, the time that the customer is willing to wait encompasses
a different number of these activities. It can occur, in multiple companies, that these steps require a long time to be
performed: recall the example of airport’ radars. The overall stages require two years, but how much the consumer is
willing to wait? The customer is willing to accept up to two years; therefore, the time he’s willing to wait corresponds,
in this case, to the total lead time of the product. However, there are cases when the clients are not willing to accept
this timing. Let’s consider the example in which a supermarket ordered Nutella from Ferrero: the supermarket wishes
to receive delivery very soon, in a few days. Therefore, Ferrero, has to produce the product in advance in order to
meet the quick delivery requested.

Now, let’s consider the different types of master production schedule, according to the type of customers’ request.

If the customer wants an immediate delivery,

the company has to carry out all activities on
the basis of a forecast and hold a stock of
Finished Products. The implication of this
model, in fact, is that all the activities have to
be carried out on the basis of a forecast (push-
approach). This is called Make to Stock (MTS)
and an example of this is fast moving customer goods.

If the customer accepts a delivery time as

long as the last production activity (final
assembly), the company has to carry out
only the first 3 activities on the basis of a
forecast (push-approach) and hold a stock of
Components and Sub-assemblies produced
internally. This is called Assemble to Order
(ATO) and an example is furniture. When you buy furniture for the kitchen, for instance, you go to the shop and
choose the design that you like and based on the space that you have at home, the furniture will be modelled
according to that request. Therefore, the customer has to wait a few weeks in order to have the furniture in line to his
needs.
Let’s consider the case in which a customer goes to IKEA and he buys a bookshelf. The customer has specific needs:
size, drawers, shelfs etc. IKEA knows all of this; therefore, the customer can find the components in order to assemble
the product that he likes. In other words, IKEA uses the assemble to order approach: it has produced in the base of a
forecast the order, and once the customer’s order arrives, the final assembly is carried out.

21
If the customer accepts a delivery time as long as the
last 2 production activities (production of components
and sub-assemblies and final assembly), the company has
to carry out only the first 2 activities on the basis of a
forecast and hold a stock of bought-out materials.
Therefore, engineering and purchasing activities are
carried out on a forecast basis, while the production of
components and final assembly are carried out on the basis of the actual order. This is called Make to Order (MTO)
and an example may be Dell Computers. Another example, may be

If the customer accepts a delivery time as long as the last 3

activities (purchasing, production of components and sub-
assemblies and final assembly), the company has to carry out
only the first activity on the basis of a forecast and hold NO
stock any material. This is called Purchase to Order (PTO)
and an example may be the production of some industrial
equipment. When a client goes to the supplier, the latter shows all the products range (engineering activity – design
of a product), which are made on the basis of a forecast on what the supplier expects customers will want. However,
the actual purchase of raw materials to conduct the order is made once the customer makes the order, so when he
chooses the product he desires from the catalogue – and of course, all the sequent activities are conducted after that
moment.

If the customer accepts a delivery time as long as all 4

activities, all of them are carried out on the basis of an
actual order and NO stock any material is held. This is
called Engineer to Order (ETO) and an example may be
construction industry or an airport radar. This occurs
when the design of a product has to be customized, so no
activities can be conducted in advance.

What are the implications of the production planning approach on the

object of the master production schedule? Depending on the production
planning approach chosen, there will an MPS referring to different
objects. If you MTS you’ll have a master production schedule will refer to
finished products, since you’ll have to produce in advance finished
products to satisfy customers. With ATO, instead, MPS will refer to
components and sub-assemblies, since they are requested in advance to
satisfy orders. Then, in MTO, the only activities carried out in advance are
engineering and procurement, so the warehouse has to have bought-out
materials. If IKEA decides to move from an ATO to MTO approach, the company has to focus on the bought-out
materials. Of course, for PTO and ETO nothing has to be produced in advance.

Example: If you were managing director of a manufacturing company, which production planning approach would you
choose between MTS and MTO? Make to Order seems more suitable, due to lower holding costs. Why does it have
lower holding costs? Let’s imagine that we’re dealing with an assemble to order manufacturing company that
produces bookcase and therefore shelves. If a customer requests a bookcase three meters high and the company can
achieve it, by putting together different shelves, since it has standard components that can be assembled to achieve
the request. On the other hand, if the company follows a make to stock approach, it has not only shelves in the
warehouse but also finished products (multiple bookcases). This could be a problem if the customer doesn’t want one

22
of the finished bookcases, therefore the company cannot satisfy the need, since it doesn’t have the bookcase desired.
On top of this, there will be some holding costs due to products not requested by anybody.
Conclusion 1: When a company has finished products in the warehouse, it is very difficult to satisfy the demand of
customers, therefore, it’ll automatically have holding costs, since finished products are not able to meet demand. In
other words, make to order approach it is more flexible, and it produces less holding costs than make to stock.

There is another reason why with MTS holding costs are higher than MTO

The first step is purchasing, then the next step is the production of
components through the materials purchased and lastly, the final assembly.
When the company buys some raw materials for the production of one unit
of the finished product, it incurs a cost of 100€. Then, when the product is
transformed, 20 more euros are spent due to extra materials, labor etc.
Lastly, after the final assembly, more money is spent, so the final unit value
of the finished product is 170 €.

Therefore, if MTS is applied the product put in the warehouse values 170, if ATO is conducted the value of the product
in the warehouse would be 120 and lastly, with MTO it would be just 100 €. This means that the holding costs would
be higher for MTS since the value of the product would be higher. Conclusion 2: this is the second reason why MTS
has higher holding costs: since the value of product is higher as well.

So, a negative implication of MTS is high holding costs, while a nice implication of MTS is definitely speed (fast
delivery).

From Production Plan to Master Production Schedule

Let’s assume that we’ve already drawn up the Demand Plan and that accordingly we have drawn up
also the Production Plan (a “Level” one).

Now, in order to develop our MPS, we have to determine the production planning approach. Let’s suppose that we’re
dealing with a MTS company, which means that when it
develops the MPS we need to refer to the final product.
Let’s assume that the product A is t-shirt, and the company
has to decide how many different types of t-shirt it will sell.
There is just one style and this model can be produced in four
different sizes. Moreover, this t-shirt can be produced either
in white or blue. Therefore, considering different
combinations between color and size, 8 different types can be
sold. These combinations in a company are called Stock
Kepping Units (SKU): is a code that is assigned to a product for the purpose of inventory management and ease of
record-keeping.
Since the approach is MTS, the MPS has to refer to each stock keeping unit.

23
From the level production plan, we move to MPS
that covers two months (totally eight weeks) and
refers to each single SKU (SKU 6 is missing – it’s a
mistake).
In January, the company is planning to produce
640 t-shirts and it split between the different
weeks and according to the eight SKU. How the
total volume is divided in four weeks? By taking
into consideration the available production
capacity of every single week. In this case, the
exact same amount is planned to be produced in
each week, meaning that the available production
capacity is always the same. However, this is not always like that: it may be different.
How the total volume is divided in the eight SKU? By considering the
demand forecasting: how many pieces for SKU 1 will be sold, how many
for SKU 2 etc. This is why there is an arrow from demand plan to MPS,
because in order to split the volume, it is necessary to conduct a demand
forecast.

Once the MPS has been completed, the MATERIAL REQUIREMENTS PLANNING (MRP)
has to be developed. Once we have decided through MPS how much to produce and
when to produce every product, it necessary to plan the production (or purchasing) of
all the components, that the finished product is made up of. In order to plan these
activities (Material Requirements Planning), we have to get some further information
about the structure of the product. This information is reported in the BILL OF
MATERIAL (BOM), which is a document that let us understand how the product is
manufactured. Through the BOM we can get two further pieces of information: – the
requirement (e.g., the quantity of Labels that is necessary to manufacture 1 unit of the finished t-shirt) – the Lead
Time (LT), i.e., the time necessary to produce 1 piece of a given component or the delivery time, in case of bought-out
materials.
In the graph on the right, we can notice that in order to obtain the t-
shirt, if we already have the label and the t-shirt without label, one
week is necessary to produce the finished product (t-shirt). Why
does it take so long? Because even though the t-shirt can be
produced in a small amount of time, it is possible that when the
order is placed, the production is busy on other orders. Therefore, this why it may require one week to be completed.
A finished product is made up by label and one t-shirt (without label): do we produce them internally? Not really,
because from the graph we don’t see any other steps before them – this means that both are made by some other
suppliers. How many labels do I need to finish a t-shirt? 1 label for 1 t-shirt and the lead time of the label is one week
– it means that from the moment in which we placed the order to the supplier to the moment in which we receive it,
we have to wait one week. On the other hand, for t-shirt
without label, the requirement is still one, but the lead time is
two weeks. Therefore, the bill of material provides all the
information regarding requirement and lead time, and it is
fundamental to structure the so-called Material Requirement
planning (on the right). It is a module of information system for
manufacturing and it makes some type of computations (when
the order has to be placed, how many units to produce etc.).
Through the MPS, we notice that on F1 we want to produce 16
units of SKU 1. The implication is that if we want to meet this
deadline, we have to compute some activities in advance:

24
going backwards, the last activity is sewing labels on t-shirt takes one week, therefore has to be placed on Jan 4. In
order to carry out this activity, both labels and t-shirts are needed: the lead time for label is one week, so the order
has to be placed on Jan 3, while for t-shirts, since they require two
weeks, they have to be requested in Jan 2 (two weeks before the
activity “sewing labels on t-shirt”).

The last two steps of the production planning process are

PURCHASING and SHOP-FLOOR PLANNING, which both comes
from MRP.
Once the orders of the bought-out materials have been planned,
the Purchasing Department can place them. Also, the
departments of the factory thanks to the MRP can know which
components have to be manufactured.

lesson 7

02.02.2021

PROCUREMENT MANAGEMENT: procurement is a department in service and manufacturing companies which is

devoted to the selection of suppliers. Some of the daily activities of this field, placing orders to suppliers, receiving
orders, satisfaction pertaining to the order received. Until a few years ago, this field wasn’t considered very important,
since the only goal was selecting the cheapest supplier (quite a fast activity, that didn’t require field). However, as
years passed, starting from the 80s companies are outsourcing more and more, especially due to products complexity
(for example, due to the amount of technology used and implemented), which require the support of multiple
companies to achieve the best quality. In the latest years, thanks to globalization, it is easier to buy materials from
cheaper countries (think about price of labor). Lastly, companies are outsourcing more due to higher flexibility. In
every industry, the innovation rate is increasing more and more, therefore the lifecycle of a product becomes shorter
shorter and sometimes, for some products, firms will need a transformation process. This is why, the production
system has to be flexible.

Vertical integration: a strategic perspective: Some relevant drivers of outsourcing decisions

• Technological innovation
• Product complexity
• Higher materials costs
• Globalized sourcing markets
• Higher flexibility

Due to outsourcing, companies have a low level of vertical integration. Let’s think about motorbike, where just 10% of
the product is made internally, while the 90% is composed by materials purchased externally or made from external
companies.
Procurement is very important, since when there is the selection of suppliers, it’s necessary to balance both prices and
quality. Therefore, this sector has to be managed in a strategic way, by using some levers that represent area of
concerns of the procurement management field.

25
Levers for strategic management of procurement refer mainly to four topics:
• Differentiated sourcing strategies: when you buy an item, you need to pursue different sourcing strategies
according to the features of items. For example, for some items is better to opt to lower prices and vice
versa.
• Vendor rating and selection: once the first step is completed, you have to select the most suitable suppliers,
so companies have to be able to conduct the vendor rating (rate that represents the level of attractiveness of
suppliers) and of course, select them.
• Price analysis: the cost of sourcing solution, which doesn’t depend only on price, but also some other issues
concerning economic outcomes.
• Managing the relationship with the supplier: if the company wants to keep on working with a specific
supplier, a specific relationship should be kept.

1)Differentiated Sourcing strategies: Product portfolio matrix (Kraljic matrix)

Depending on the features of the product you have to buy,

you have to set different aims for the procurement
management. In the case of the motorbike, you have
optimize quality. When multiple materials are bought, it’s
important to cluster them in different groups that have
specific resources strategies – of course it’s not possible to
do it for each single product, that’s why it’s fundamental to
cluster them in different categories. Through the Kraljic
matrix, it is possible to divide bought out materials in four
groups, each one having different sourcing strategies. The
variables in order to find the most suitable sourcing
strategies are:
1) Profit impact: high if companies spend a lot of money on a specific item (i.e., t-shirt, 50/60% of the value of
the product is spent to buy cloth, therefore t-shirts have a great profit impact). It can also occur that even if
the money spent to buy a specific item is low, the profit impact is still high when the product is able to drive
customer satisfaction. For example, let’s consider a Moncler coat and suppose that the customer is satisfied
when he purchased it, but after a few days he notices that the zip doesn’t work properly. Of course, the
customer will be no longer satisfied, therefore the zip is an item that has a great influence on customer
satisfaction, but at the same time it’s very cheap. Therefore, even if the company didn’t spend a lot on the
zip, since it drives customer satisfaction, it has a high profit impact.
2) Supply risk: high supply risk when it’s hard to identify the supplier that meets company’s need. This occur
especially when we’re in monopoly and oligopoly, there is one player or just a few players. On the other
hand, when there are a lot of suppliers there may still be cases with a high level of supply risk; for example,
when the level of quality for the majority of suppliers is low or unstable. Therefore, even though there are
potentially a lot of suppliers, it’s not easy to find a suitable one. Low supply risk happens when there are
several suppliers that are able to satisfy the need of the client – this usually occurs when the product is a
commodity, therefore it is produced by multiple players.
Now that the two variables are clear, it’s fundamental to describe the characteristics of materials that arise from
these two variables:
- Noncritical products (low profit impact and low supply risk). Therefore, they are no expensive, they do not
drive customer satisfactions and are usually commodities. These items are called noncritical and efficiency is
the strategy that should be pursued. In this we find stationary products like pens, rubbers and pencils; the
aim for this category is cutting costs, which can be achieved by choosing the less expensive supplier and/or
by buying large quantities to save money. However, it’s important when companies decide to save money on
noncritical products, to make the procurement phase as fast as possible: because if a company takes 1 month
to select the cheapest supply, and the person who has to make this decision is paid weekly several monies, at

26
 the end of the selection phase, the savings won’t be that convenient. Therefore, it’s important to find a
cheap supplier, but not the cheapest one, in order to keep the process very fast and simple.
- Bottleneck products (low profit impact and high supply risk). Therefore, they don’t drive customer
satisfaction and are not expensive, but they are in a supply market quite problematic. A typical bottleneck
product is packaging. An example, is a fashion brand that produces different leather shoes of different colors
and decided to produce boxes of the same color of the shoes (if shoes are red, the box would be red as well
etc.): this is a bottleneck product, since boxes are not expensive, they don’t drive customer satisfaction but if
a company wants boxes of different colors, it has to establish a partnership with a specific supplier (common
suppliers produce box of basic colors, it’s hard to find suppliers that produce boxes of those colors) –
therefore, in the supply market, just one will be able to satisfy company’ needs. This is very risky, because the
supplier is not easy to be replaced in case of need. How to solve this situation? The strategy to adopt is a two
steps strategy. The first step is to increase and accumulate stock, in order to be protected in case of need (for
example, the supplier’s machine is broken) and secondly, the material has to be changed. The implication is
that it’s possible to move the product from a bottleneck situation to a safer one.
- Leverage product (high profit impact and low supply risk). This is the case of cloths and shirts, since they
drive high customer satisfaction, companies spend a lot of money on cloth but at the same time, cloth is a
commodity, so there are multiple suppliers. The strategy that should be pursued is the so-called competition
strategy: since the product is a commodity, the quality can be taken for granted, companies can choose on
the basis of the price. In order to find the cheapest one, a buyer should be employed in order to look for the
cheapest supplier – if the final savings is huge. Generally, due to competition, is quite easy to find the
cheapest supplier.
- Strategic materials (high profit impact and high supply risk). For instance, by considering toothpaste, what
does drive customer satisfaction is the tube, which is a demonstration of quality, and on top of this
companies spend a lot of money on this item. Moreover, there are a lot of suppliers but just a few provide
high quality tubes. The strategy to implement is a partnership: once the company finds the supplier perfect
for its aim, it should set a partnership that lasts for
years (it’s a real contract).

Dynamic aspects of Kraljic matrix

All materials and services bought from suppliers, are divided in
the different categories. Therefore, Kraljic matrix is a picture of
all the materials bought and the strategies to apply. However,
this matrix can be used also to have some suggestions for some
repositioning of the materials inside the matrix. The first one is
that if we have bottleneck products the suggestion is to create
some stocks in the short term, but in the mid-term an
alternative item should be found, such that when this
alternative product is found the original one turns to a
noncritical product. The second repositioning is the area of
noncritical product to leverage product. For example, in universities there are multiple stationary products, which are
usually bought from multiple suppliers. However, if instead the university decides to purchase all products from a
unique one, a total amount spent on stationary products is not relevant anymore – when this occur, the products will
be moved in leverage products, so the company should no longer look for efficiency but for competition (supplier at
the lowest price). The third repositioning is from leverage product to strategic materials. For instance, Sony produces
camera, computers and several electronic devices, but in the 80s the most common product was definitely for
Walkman, since it was able to reduce the size of these devices. However, technologies changed, Walkman
disappeared, so Sony had to produce new product lines to replace this item and created cameras with lenses from
Zise: if Zise wouldn’t be able to produce lenses anymore, Sony would be in a very difficult situation – therefore lenses
are strategic materials, since can be purchased by just one supplier.

27
The last repositioning occurs for instance when Sony, after acquiring a lot of fame for its cameras, decides to purchase
lenses for multiple suppliers and from a partnership strategy moves to a competition one (therefore there is a
movement in the leverage product field).

SUPERSKIING CASE STUDY

Superskiing produces the following types of products:

• Hardware for climbing (climbers, helmets etc.);

• Clothes, bags and other accessories for outdoor activities on the mountains;
• Safety clothes and equipment for worker.

The Procurement Manager is wondering whether the procurement strategies of the company are correctly defined
and, for this purpose, he wants to draw up the Kraljic matrix. Relevant information in this regard is reported below:

28
 - Draw the Kraljic matrix
- What do you think about the positioning?

In order to place the materials assessment on Profit Impact and Supply risk have to be made.
Remember, that when we analyse profit impact we both have to consider whether the company spends a lot of
money on that material and if the materials drive customer satisfaction.

Profit Impact: % of purchasing cost – do we spend a lot of money on a given material?

We can notice that for Standard Steel, the company spends a lot of money (26%), so it should be placed on the upper
part of the matrix, while Others (5%) in the lowest part. However, Insulating cloth is an item with an intermediate
value (13%) – therefore, it’s fundamental to identify a threshold through the so-called Pareto analysis: 80% of
phenomenon can be explained 20% of causes. Here, the phenomenon is the total amount of money that the company
spends, while the causes are the materials.
How is it possible to identify that 20%? In order to do that, it’s necessary to rank all the items from the one with the
highest impact on the purchasing costs to the lowest one (table 2, after the arrow). The next step is to compute the
cumulative percentage and through this step it is clear that 80% of the phenomenon can be explained by the first four

29
items: STDS, SPCS, PLAM and ICL, even though they are not the real 20%. These items would be placed in the upper
parts, while the other ones in the lowest one.
However, also the ability to drive customer satisfaction has to be considered: it could be that for some items that do
not represent a high expense, they may drive customer
satisfaction and vice versa.

In order to make an assessment on customer

satisfaction, some quality observations have to be
made. If there is at least one yes, then there is high
profit impact.
By considering both aspects, the profit impact can be
established.
In this case, it’s necessary to make a distinction between
general accessories and zip.

Supply Risk

This assessment is quite easy, since it depends on the number of

suppliers. Of course, if materials can be bought from several
suppliers, the supply risk is low, while if there are just a few
suppliers, the risk is high.

Kraljic Matrix

By putting together all this information, the matrix can be

formed. The area of the bottleneck products is empty, which is
very good. However, apart from that, some movements can be
suggested. The number of products in strategic materials is too
many, because it means that the company has too many
partnerships, which are risky (due to the exclusive relationship
with a supplier) and tough to manage. Plastic materials are
usually commodities, but in this case study, the company asked
for some specific technical requirements – if these
requirements/characteristics are not so important, these
products could be moved in the leverage product category,
where plenty of suppliers can be found and reduces costs thanks to leverage.

Number of suppliers

In this case study, there is a problem concerning the number of

suppliers.
- Sole sourcing: company purchases from one supplier because there is no alternative. For example, in the 80s
just Telecom was available if people desired to purchase telephones.
- Single sourcing: company purchases from one supplier, even if there are multiple alternatives. For example,
nowadays we can purchase telephone services from Telecom even though there exist multiple options.
The risk with this sourcing is that the company relies just on one supplier, and this can be solved through
second sourcing.
- Second sourcing: when there are two suppliers, the main one and the secondary one. The company can
decide to purchase 90% of items from the most important one, and 10% of products from the second one.
This is good, since if something happens to the first supplier, the second one is available, and in that case,
he’s able to increase the volume of products to supply.

30
 - Parallel sourcing: used by big companies which purchase big volumes (i.e., car manufacturing). For example,
Fiat, which produces multiple auto vehicles, and of course for each category the company needs lights. By
using parallel sourcing, the company will purchase all the lights of one type of car from one supplier and so
one (one supplier for 500’s lights, one supplier for Punto’s lights etc.). However, these suppliers know that
they work on parallel with other suppliers. Through this method, it is possible to reduce costs due to the
high-volume purchases, and the supply risk is also very low.
- Multiple sourcing: companies buy the same material from different suppliers. There are not a lot of benefits:
few quantities of purchases, multiple suppliers to manage etc. It’s a situation that can happen, but it’s not
really a strategy to follow.

lesson 8
05.02.2021

2)Vendor rating and pre-selection

Once the differentiated sourcing strategy has been identified, the next step is to identify the most appropriate
supplier: how is it possible to achieve this goal? A process has to be conducted, the so-called vendor rating and pre-
selection, which consists of three steps:
1) Market Research: it consists in understanding how the market is organized, so how many players you can find
in a market (just a few big players, a lot of small players) and how to manage them. It’s very important
because if you’re dealing in a market highly concentrated, with just a few players, my bargaining power will
be very low, so it would be difficult to find a good supplier. On other hand, if there are a lot of suppliers my
bargaining will be higher, but there is a draw-back in the situation: since if suppliers are too small, maybe
they will be artisans, so not in the condition to work with me, if I’m a big company. If the suppliers are too
small, maybe they are not able to satisfy my demand and request.
Once you get in touch with supplier you choose and collect information regarding them: like, do they have
some specific certifications (like ISO certification of quality management, certifications regarding the
environmental system). Then, in order to identify from a group of selected suppliers, the perfect one, the
next step has to be performed.
2) Vendor Rating and Selection: process in which the company gets in touch with the selected suppliers and ask
them what are the conditions on which they could sell items. The conditions that the company asks for are
related to three possible performances: 1) price, 2) quality level, 3) time (speed of delivery, dependability of
delivery etc.) – quite important if you need quick delivery in some companies.

On the basis of these three performances, the company computes the vendor rate for each single supplier.
The vendor rate is a weighted average of quality, time and price; it is an average since there are some weights
used to express the level of important of quality, time and price in the final decision. The numbers obtained
are used to rank the suppliers, in order to select the highest in the ranking. There are a couple of complexities
however: 1) weights (𝛼, 𝛽, 𝛾): how to select them? This is a managerial decision. For example, consider that
you want to buy a commodity, quality is taken for granted, therefore if you’re dealing with an item like this,
supplier would be probably selected on the base of price, therefore 𝛾 = 50%, then quality 𝛽=30% and time
which is the least important 𝛼 = 20%. In other cases, of course, percentages according to importance, would
be different. 2) units of measurement: for quality would be the percentage, since it is usually expressed in
terms of scrap rate, time is expressed in number of days (usually for delivery) and price is expressed in €. How
to compute an average between those? A system has to be introduced to transform this in a unique unit of
measurement. How to do that? Let’s consider the following example.

31
Example: we decided to buy a product, for
which weights assigned are the following:
- Cost: 20%
- Quality: 40%
- Dependability: 40%
A way to transform these in the same unit of
measurement is to establish a scale (like 1 to 3
scale), where 1 is a low performance, 2 is a
medium performance and 3 is a high
performance. We buy a product which is quite
expensive, the performance of the supplier is low (1),
then I know if the price is between 1,1€ and 1,4€ it’s
quite okay, so the grade would be 2 (medium
performance), while if he asks for a price less than 1,1
€ would be cheap, so a high performance (3). The
same thing would be done for quality and for
dependability as well.
On the basis of these numbers, it’s possible to
compute the vendor rate of suppliers. In the table on
the right, for each supplier there is a performance
score, for each single category: cost, quality and price.

From these scores, the vendor rate can be

computed, by of course multiplying each score for
the weights. It’s clear, through the results
obtained, that the best supplier is the supplier
number 2, since he is the one with the highest
vendor rate.

3) Ex-post Vendor Assessment: once the best supplier has been selected, now the company starts purchasing
products from supplier and once the collaboration has been concluded (suppose after 1 year), ex-post
assessment can be conducted. For example, in the previous example, supplier 2 declared that he would be
able to perform 98% in dependability and 1% in quality; what it’s possible to do, is to compute again the
vendor rate based on the actual data, after the performance. It could be that it would be higher or lower –
for example, suppose that the actual values are 5% for quality and dependability 94%, it means that quality
falls in the medium performance (2) and dependability in the low one (1), therefore the vendor rate would be
lower. If this happens, since the supplier 2 vendor rate is lower than 2 ( so lower than supplier 1), the
company has to make a decision: asks to supplier 2 for an improvement (by also using some penalties – if
he’s not able to meet the requirement, he’s going to be replaced by another supplier).

CASE STUDY – MY T-SHIRT

• T-shirt A is a trendy item, require a fast delivery (delivered in no more than 2 weeks – even though the
average 3.5 weeks). A is longer and more complex manufacturing processes, which involve on average a unit
purchasing cost of 12 €. In the recent years, MyTS has been able to negotiate for a similar product a unit
purchasing cost of 10 €.
• T-shirt B is a carry-over product, which is sold in all collections. much simpler and, considering its features, in
this industry it is possible to procure it at an average cost of 8 €, but MyTS has been able to recently buy a
similar product at a unit cost of 7 €.

32
Also, quality should be considered (average 8%) – My T-shirt wants high quality standards (scrap rate 4%). A further
element to consider in the vendor selection process is the environmental sustainability of the subcontractors. The
estimated carbon footprint of a t-shirt along its lifecycle is 10,75 kg of CO2e, 3 kg of which result from the production
process, 1/3 due to the heating processes and 2/3 to the consumption of electricity. In the vendor selection process,
MyTS prefers those subcontractors that are striving to improve their overall environmental performance.

Request: On the basis of the available information, MyTS must decide which subcontractor is the most appropriate
one for products A and B.
Solution:

Which weights should My T-shirt use?

T-shirt A: Delivery is at the first place, then quality, then sustainability and lastly price.
T-shirt B: Price is first, then at the same level quality and sustainability, and lastly delivery time.
It’s important in this step to identify priorities – of course different percentages may be correct, but priorities have to
be respected.

Can we compute the vendor rates?

Starting with A: Since there different measurement systems, a scale should be constructed.

12 € is the average of the industry, if it is higher would be bad, therefore 1 (low performance). 10€ is an average value,
on the other hand, therefore 2 (medium performance), and of course if it is lower than 10€ that would be perfect 3
(high performance). For the delivery time we know there is a threshold: higher than 3,5 weeks would be really bad,
while below 2 it would be great. The average, as expressed in the text, is between 2-3.5 weeks. The scrap rate follows
the same procedure, and from the same reasoning we identified the three-performance level. Finally, environmental
sustainability doesn’t have a very specific information, but we can deduct that more certifications bring higher level of
performance.

33
Therefore, the vendor rates for product A are the following, on the right.

For the product A, therefore, Alpha would be the best subcontractor, since it’s higher than Beta’s one.

The same long procedure, can be conducted for t-shirt B, and what we obtain is:

For product B, the best subcontractor would be Beta.

What happens if we change weights? The results of the vendor rates would be different, and consequently also the
choice of the supplier would change.

3) Price analysis
When you buy a product, sometimes you don’t have to give a lot of importance to price, since quality or time may be
more important: this doesn’t mean, however, that price should be overlooked. The real challenge is to understand
how much it costs to purchase from a supplier. The cost doesn’t always correspond to the unit price of the product
purchased from the supplier, but also other costs should be considered:
- Transport
- Packaging
- Maintenance: especially, when an equipment has a long life cycle, maintenance activities could be needed.

34
 - Holding costs
- Quality level
- Etc.
Therefore, unit cost has to be added also to these costs, or the costs incurred in the specific case à Total cost of
ownership: how much does it cost buy and own that product?

The total cost of ownership, in this case, is lower from the supplier that charges the higher unit price.

4) Manage the relationship with supplier

To promote its own image with the client and strengthen this relationship, a company can undertake actions in the
following areas:

• Technical activities: sometimes the client is a big company compared to the supplier, and so he can explain to
the supplier how to improve, by for example organizing training programs for free or by arranging a
consultancy project. The supplier will improve its level of competency basically for free; of course, it can
occur, the supplier by improving, will incur a lower level of cost and the client can request a discount. Still, it
is convenient to the supplier, since he’s improving his abilities.
• Financial activities: the client wants to be a considered a good client by either paying on time or before the
deadline. The supplier would appreciate this, and he’ll be able to plan his future revenues – this could be
really useful for the financial plan.
• Planning activities: if you’re the supplier, you receive orders from the clients and you hope that you have
enough stocks to satisfy the client. You can improve, if your client can inform in advance the supplier with the
pieces that will be requested in the future. There are multiple companies, which share with their suppliers
their production plan in advance, so that the supplier can be informed properly about the future requests of
clients and prepare the production. The supplier can avoid the risks of overstock or stock-out and can avoid
forecasting.

PROCUREMENT AND SUSTAINABILITY

Nowadays, companies are more and amore aware that not only economic success is
important when running a company, but also the safeguards of people and environment.
On the basis of this analysis, the 3P approach has been created, where the 3P stands for
profit, planet and people Profit has to be generated when running a company for
shareholders, but also the focus has to be on social and environment aspects. Only if
these elements are combined, the company will be run in a proper way.
Planet, recalls the topic of sustainability, how can be appointed? First of all, let’s consider
the following approach, which is linked to meet sustainability.

35
The lifecycle assessment is an approach for
running a company, in order to assess the topic
of sustainability. The life cycle relates to the
life cycle of a product which consists of:
1) Design of a product
2) Procure the materials to run
production activities
3) Produce the product: transform raw
materials into finished ones
4) Distribute: take care of transportations that may be highly pollutive
5) Use: the usage of the product – a lot of emissions are used usually during this phase (like t-shirts, when you
wash them, you’ll use some waters, detergents that pollute the water, iron etc.)
6) End of life: when the product has terminated its life and when you get rid of it, it may have an impact on the
environment.
When you want to reduce the environmental impact with your company, you have to identify firstly how many
emissions you generate with each step and secondly the step that pollutes the most. For example, if the production is
the most polluting one, your focus would be on that activity, in order to reduce CO2 emissions.
In each step of the lifecycle of the product you can adopt some procedures that can be helpful in improving the level
of economic, environmental and social sustainability.

Today we’re going to focus on procurement and how the level of environmental sustainability in particular for
procurement activity.

The role of purchasing in delivering sustainable solutions

In order to find the tools to used, it’s good to identify the main steps undertaken by a company in order to buy a
product.
1) We have to identify needs and define
specification; therefore, the company
decides the product it wants to buy and
therefore, the need. For instance, if I want
to produce shoes made of leather, I need
leather to be purchased. Of course, in this step some specifications are made (we want to buy a certain type
of leather, due to durability), in order to have a detailed guideline of what the company wants to purchase.
2) Vendor pre-selection: it is the activity regarded to the identification of the supplier from whom to buy in the
future; a market research has to be conducted to identify a set of potential suppliers from whom to buy.
3) Tender evaluation and vendor selection: vendor rating process is conducted on the set of potential
suppliers, in order to select the best one.
4) Vendor control and contract management: once the supplier is selected, the company starts buying from
that supplier and after a certain amount of time, it is time to control the actual performanceof the supplier,
whether to understand if he’s doing well or not.
In each step, some solutions can be taken to bring the company towards the achievement of sustainability. Let’s start
with the first one:

Stage 1 – Identification of needs and defining specifications

The possible solutions that can be undertaken are:

36
a) Reducing consumption: when an item is purchased, a certain volume has to be chosen. This total amount can
be affected by the scrap date. Let’s consider the following example:

If in a year 10 good quality pieces have to be purchased, if I purchase from supplier 1 all pieces pass the quality
test control, while with supplier number 2, every 10 pieces there is 1 scrap. Therefore, to have 10 good quality
pieces, 11 have to be purchased. Consequently, more euros have to be spent.
The scrap rate, therefore, can make a difference both from an economic and environment point of view: if I don’t
consider quality, I waste some resources and materials.
Therefore, the opportunity to address overall consumption should not be overlooked, by:
a. Reduce total quantities (e.g. giving priority to suppliers with a lower scrap rate)
b. Reduce excessive stock (which generally results in
higher obsolescence rate)
b) Challenging repeat purchase: do you really need to buy that
specific material?
Suppose for example that a company has always bought
plastic pallets for transportation and warehouse activities, for
years and years: it doesn’t challenge its purchase, so it doesn’t
consider purchasing any alternative products, from the environmental viewpoint. For example, plastic pallets
could be substitutes with eco pallets, that perform the same function. Sometimes, it may be interesting to
consider alternative products.
c) Identifying alternative solutions: the opportunity to buy a service that delivers a function rather than a
product should be considered
~ E.g., rental and/or car sharing: suppose that a company needs some cars for its activity, and it
purchases them for the period needed, and once it doesn’t need it anymore or it has to replace it
since it is updated, of course it causes a waste (raw materials, energy, production activities that
embody CO2 emissions etc.). However, it’s possible to cope with the same need, by using car sharing
or by renting cars, just for the hours that the company needs cars.
d) Specifying greener products: the company can set environmental requirements for the selection of bought
materials. E.g.:
~ Procuring greener pieces of equipment
~ Prefer renewable energies
~ Avoid hazardous substances
~ Avoid ozone-depleting substances
~ Prefer certified products/suppliers (ISO140001, EMAS)
~ Etc.
These help to increase the level of sustainability in the procurement phase.
e) Product design and development: companies could encourage the use of techniques as Design for
Environment or, generally speaking, design approaches that are more eco-friendly.
For example, shirts made by Patagonia are made by recyclable materials and they are 100% recyclable as
well. By collaborating with suppliers, it’s important to understand whether they are up for these sustainable
implementations and in the design phase, therefore, it’s important to use techniques more eco-friendly. If
the client is a big client (i.e car industry) he’s in the position to teach the supplier how to implement these
techniques and the client in a certain way, takes care of the designing phase.

37
Stage 2 – Vendor pre-selection

a) Pre-selection: Several techniques can be used at this stage to identify the best possible supplier:
• supplier questionnaires
• supplier visits: to be sure that the supplier has essential features to be accredited.
The output of this activity is the accreditation of the supplier.
Assessment questionnaires example – addressed to the supplier, to understand whether he’s good from the
environmental point of view.

Other questions may regard standards and certifications:

• ISO14000 – ISO 14000 is a series of international standards on environmental management. It provides a

framework for the development of an environmental management system and the supporting audit
programme – ISO 14001 is the corner stone standard of the ISO 14000 series. It specifies a framework of
control for an Environmental Management System against which an organization can be certified by a third
party. This standard is applicable to any organization that wishes to:
o implement, maintain and improve an environmental management system
o assure itself of its conformance with its own stated environmental policy (those policy commitments
of course must be made)
o demonstrate conformance
o ensure compliance with environmental laws and regulations
o seek certification of its environmental management system by an external third party organization
o make a self-determination of conformance
• EMAS (Eco-Management and Audit Scheme) – The EU Eco-Management and Audit Scheme (EMAS) is a
management tool for companies and other organizations to evaluate, report and improve their
environmental performance.
• ISO 50001 – Provides requirements for organizations in all sectors that want to use energy more efficiently,
through the development of an energy management system (EnMS).
• ISO 50001:2011 provides a framework of requirements for organizations to:
o Develop a policy for more efficient use of energy
o Fix targets and objectives to meet the policy
o Use data to better understand and make decisions about energy use
o Measure the results
o Review how well the policy works, and
o Continually improve energy management
• SA 8000 (Social Accountability) – Provides requirements for the sustainable implementation of the principles
concerning: child labor, forced and compulsory labor, health and safety, freedom of association and right to
collective bargaining, discrimination, disciplinary practices, working hours, remuneration.

38
 • OHSAS 18001 (Occupational Health and Safety Assessment Series) – Is an international occupational health
and safety management system specification. It is intended to help an organization to control occupational
health and safety risks
Lastly, a company can also evaluates:
• Code of Conduct – Is a set of rules that describe the responsibilities and the practice of an organization to
pursue a sustainable and ethical conduct.
Example: the Apple Supplier Code of Conduct
- Apple is committed to ensuring that working conditions in Apple’s supply chain are safe, that workers are
treated with respect and dignity, and that manufacturing processes are environmentally responsible.
- Apple requires that Suppliers implement this Code.

Stage 3 – Tender evaluation and vendor selection

• At this stage, the company compares all bids by assessing their
attractiveness against pre-determined parameters. Such parameters
generally concern:
o Cost ($/pc)
o Quality (e.g., scrap rate)
o Time Performance (speed and dependability of delivery)
o Social & Environmental features
• The output of this process is the Vendor Rate assigned to each supplier/bid, which finally drives the supplier
selection.

Example: Disney
Disney expressed how subcontractors for t-shirts were selected. Initially, there were three subcontractors that were
compared according to time and costs: one in the US, one in Mexico and one in Vietnam.
The process was so:
1) First cost: First, consider the unit cost to pay to the subcontractor: on the basis of this, the Vietnam
subcontractor was considered the most convenient.
2) Landed cost: Costs that take into account custom duties. By looking at the landed costs, the cheapest was the
Mexican one.
3) Delivered cost: the Mexican solution was considered the best one.
4) Transit time: again, the Mexican and American solution were the best one, but of course the Mexican was
the best one, since he was the cheapest.
This was quite controversial: it’s quite frequent that companies from the US outsource production to Mexican
companies. These Mexican factories that produce for American clients are placed next to the border, in order to
reduce the delivery time. However, these factories are in the middle of nothing, basically in the desert, which is very
dangerous, also for workers that have to go there on a daily basis (many sources of violence can be seen). Female
workers are very frequently raped during their working time. This could be very dangerous for the brand image of
Disney. The shareholder value, in fact, it’s also affected by reputation, which is a key element to consider. Therefore,
Disney decided to change its vendor rating system, and it introduced some new elements regarding for example, the
compliance with international labor standard laws and environmental initiatives. All the elements are scored, to a 1 to
5 system, weights are not percentages, but they are expressed in a 1 to 3 scores. The highest possible score is 100-

39
 lesson 9

Stage 4 - Vendor Control and Contract Management

At this stage, the company has to manage the relationship with the suppliers for the life-time of the contract. During
this period, the company:
• Periodically measures the ex-post performance of the supplier
• Can set more challenging goals with the supplier, in pursuit of a continuous improvement approach
• Promotes continuous improvement through workshops and conferences with suppliers
In this phase, the company can evaluate whether to keep working with the supplier or not. It’s important to consider
that when a company manages the relationship with the supplier, it can contribute to supplier’s improvement
concerning environmental issues.
For example, Illy Caffè works a lot with suppliers of coffee (farmers in southern America – small companies, don’t have
a lot of managing skills and information regarding sustainability). Illy is supporting them a lot, by also creating the
University of Coffee, that provides training programs for these farmers. This is a way to have a social responsible
behavior, and support suppliers to make them work better.

lesson 9
19.02.2021

HERBERT KOTZAB

A supply chain logistics decision framework

Drivers of Supply Chain Performance

• Facilities: where are the places where I produce, assemble or store products or inventory – this is about
location. Production sites and storage sites.

40
 • Inventory: raw materials, WIP, finished goods within a supply chain. Inventory policies.
• Transportation: how do we move inventory from one place to the other, what is the transport mean that we
use. Combinations of transportation modes and routes.
All these three are interrelated, which means, that if the company faces a problem in facilities, it will impact both
inventory and transportation.
• Information: if I have information pertaining to just facilities, inventory and transportation I can take good
decisions. Information refers just to the logistics and supply fields. It is data and analysis regarding inventory,
transportation, facilities throughout the supply chain. It is potentially the biggest driver of supply chain
performance.
It’s not possible to develop a supply chain strategy without a competitive strategy. Therefore, strategic fit means that
both the competitive and supply chain strategies have aligned goals. It refers to consistency between the customer
priorities that the competitive strategy hopes to satisfy and the supply chain capabilities that the supply chain strategy
aims to build. The issue of achieving strategic fit is a key consideration during the supply chain strategy or design
phase. Different companies adapt different competitive model.
The priorities to pursue when designing and managing a SC depend on the nature of the product, which can be
distinguished into two main categories:
- Functional products: generally commodities, that are characterized by a stable demand pattern and,
therefore, their sales can be easily forecasted (e.g., detergents are functional products). In this case product
innovation is not common.
- Innovative products: with high innovation rates, which are subjected to a quick obsolescence due to fashion
trends or technological change (e.g., Zara products, or all fashion products, other industry concentrate on
innovative products are smartphone and pc companies).
Finding the right fit
There are two main kinds of products with some specific features and
depending on the nature of the products we have to decide the supply chain:
if you gave completely innovative products it means that you have the so-
called Responsive Supply Chain. On the opposite, if you produce and sell
functional products you need to have an Efficient Supply Chain. So, there is a
match between the responsive supply chain and innovative products- and
between. Efficient supply chain and functional products. Responsive supply
chain because they react to customer demand, they respond to market
needs. It’s the supply chain of Zara because all the decision they taken are
aimed to guarantee the availability of the products. This is the challenging in
the fashion field. The main important thing is to sell products under any condition and try not to accumulate too many
things in the warehouse. Efficient supply chain responds to speculated/forecasted rather than actual demand. It’s able
to minimize the logistic costs. Functional products are characterized by low margin, stable demand function and low
innovation (e.g., detergents). In this case is not really important to sell products under any conditions. According to
Fisher, the two supply chain typologies have the following features:
Þ Reactive supply chain: it is able to promptly respond to market needs, ensuring a quick availability of the
product, thus minimizing stock-outs, even though this can involve some production and distribution
inefficiencies.
Þ Efficient supply chain: it is able to minimize the total logistic cost through an appropriate organization of
production and distribution activities.

41
Modularity: is the degree to which a system's components may be separated and recombined, often with the benefit
of flexibility. It occurs when brands have modules that can assemble in different ways; with this, companies allow
customization.

Long-medium-and short-term decisions

Þ Supply Chain Design: how is the supply chain structured and what processes are performed at each stage?
Þ Supply Chain Planning: how are markets supplied?
Þ Supply Chain Operation: how is the individual customer order fulfilled?

Organizing the supply chain by identifying the Order Decoupling Point

In our chain we have points (the ones in the line) called order decoupling points: where is dependent demand
meeting independent demand. Therefore, am I able to wait the finalization of product until I receive an order
(customer driven) or do I have to produce in the expectation of the demand, and then I move the finished product to
the final destination and there, customers are able to buy finished products (forecast driven).

CAPACITY MANAGEMENT AND S&OP

What is capacity? The capacity of a worker, machine, work center, plant or organization to produce output per time
period.
Decisions to be made:
- How is capacity measured?
- What factors affect capacity?
- What is the impact of the supply chain on the organization’s effective capacity?

42
How does the supply chain affect the workers’ capacity?
Example: Capacity for an assembly plant with three assembly lines and a maximum of two 8-hour shifts per day:
Capacity = (800 units per line per shift) (number of lines) (number of shifts)
Is 800 the maximum or the average?
There are some factors to consider:
• Controllable factors: number of shifts, number of lines,
training of employees, etc.
• Uncontrollable factors: supplier problems, conformance
quality, late or on time
How can I fit the capacity to the demand? In this case, demand is
always increasing over time.
Capacity expansion strategies
v Lead capacity strategy: in this case, demand is always
increasing over time and capacity is added in anticipation
of demand increasing (usually companies reason like that).
v Lag capacity strategy: capacity is added only after
demand has materialized (usually public sector reasons
like that – it is always behind getting larger).
v Average capacity strategy: capacity is expanded to
coincide with average expected demand. This is what
probably Zara does.
Make or buy decisions are decisions regarding the choice between insourcing and outsourcing, where:
Þ Insourcing: inputs from within the firm (production inside)
Þ Outsourcing: inputs from outside the firm (production outside)

Insourcing/outsourcing decision process

Sales and Operations Planning – S&OP: A process to develop tactical plans by integrating marketing plans for new
and existing products with the management of the supply chain. Brings together all the plans for the business into one
integrated set of plans (sales and operations planning).

43
Top-down planning process:
1) Sales forecast
2) How much it will take to produce? How many working hours per
day I need? How much does it cost to produce a unit? How much
does it cost to outsource? What if I reduce/increase my
workforce – how much does it cost?
3) Planning different ways to produce the quantity desired.

Demand forecasting: laws of forecast

• Forecasts are almost always wrong by some amount (but they are still useful): sometimes they are more
wrong, sometimes less – forecast is always a guess.
• Forecasts for the near term tend to be more accurate.
• Forecasts for groups of products or services tend to be more accurate – since there is a consolidation
• Forecasts are no substitute for calculated values.

Designing a forecast process

Important input to forecasts

All this information is available, it represents the expert knowledge.

Forecast methods depend on data available and time horizon. Predict the systematic component of demand and
estimate the random component

44
 usually regression

Qualitative methods are used when there are no quantitative data and when there are long/medium range plannings.
Quantitative methods like time series, simulation and causal relationship are used for short term, while econometric
methods for more long-term evaluation.

What does a time series consist of? In this timeline, there are seasonal
peaks (up and down line) there is volatile demand. The demand is
following seasonality and a trend which is increasing. Random variation,
on the other hands, are situations that cannot be predicted and can
occur time to time.

Components of a time series:

• Randomness
o Unpredictable movement from one time period to the next.
• Trend
o Long-term movement up or down in a time series.
• Seasonality
o A repeated pattern of spikes or drops in a time series associated with certain times of the year.

What about including new information? It occurs, especially when you do forecast in a company, new information
occurs continuously.
Þ Static approaches
o the estimates of level, trend and seasonality do not vary as new demand is observed →
Decomposition of a time series. This is applied when new information is not added.
Þ Adaptive approaches
o the estimates of level, trend and seasonality are adjusted after each demand observation → Moving
averages, simple exponential smoothing, Holt’s model, Winter’s model. This is applied when new
information is considered and added to forecasting.

General overview of forecasting methods for time series regular demand is

different from
irregular demand,
cerca differenza

45
To forecast either the moving average or new forecast method
can be used.
Moving average: I have a time series, with n observations, and I
decide upon a time window and compare it with a previous one,
take the average of this and this is to be the estimate for the next
time period.
vedi foto e
New forecast (SEM) – simple exponential mothing
slides numeri
Forecast error: last period’s actual demand – last period’s
esercizi
forecast. It’s the deviation about what we thought it would
diversi
happen and what really happened.
If the smoothing constant is close to 0, the forecast reacts slowly, while if it close to 1 the forecast varies very nervously.
Forecasting exercise 1:
Use a four-day moving average to forecast the Friday demand.

(140.4 + 160.0 + 139.9 + 123.6)

𝐹456789 = = 139.725
4

Let’s say the actual sales for a given day totaled 115 pounds. while
the forecast for that day was 110 pounds. With a smoothing
constant of 0.10. what is the next day’s forecast?
110 + 0.10 × (115 − 110) = 110.5

In this time series we notice an increasing trend and a seasonality,

since Friday is always the lowest and Saturday the highest.

How to put the three systematic components S in a

static model together? (level-trend-seasonality)
These three methods, all put in relation the three
components. In the real life setting all three
approaches are tried.

Four-step procedure for adjusting forecasts for

seasonality
1. For each of the demand values in the time series, calculate the corresponding forecast using the unadjusted
forecast model – for each value of time series, we’re going to find an unadjusted forecast model, through a
regression.
2. For each demand value, calculate the ratio (seasonality factor or seasonality index) (Demand/Forecast). If the
ratio is less than 1, then the forecast model over-forecasted; if it is greater than 1, then the model under-
forecasted. So, we either forecasted too much, or too less.
3. I have to check how many observation periods do I have in my calculation. If the time series covers multiple
years, take the average (Demand/Forecast) for corresponding months or quarters to derive the seasonal
index. Otherwise use (Demand/Forecast) calculated in Step 2 as the seasonal index.

46
 4. Multiply the unadjusted forecast by the seasonal index to get the seasonally adjusted forecast value.

Unadjusted forecast model

The red line is the demand, green line is the
regression line and it represents the best fits
of values. The R2 tells me how much of a
variance is explained by this function. The
closer is to 1, the better it is: it’s able to
explain the system perfectly.

Customers= 122.81+27.02*Day, where 122.81

is the level that I sale no matter what day is it
and depending on the day I can sale
additionally 27.02 euros (day 1 = 27.02, day 2= 54.04).

Demand/Forecast Ratio

Consideration of seasonality

47
 Adjusted forecast model

lesson 10

23.02.2021
lesson 10
Forecast accuracy
• How do we know if forecast that we did is wrong or
not? Is the calculation correct or not? The best
calculation is the one with the lowest forecast error
• We take our estimations, we go back and time and
we modeled the past and check out which model provides the result closest as possible to what really
happened.

Measures of forecast error

Forecast error is the difference between what we’ve

forecasted and what really happened.

• Mean Absolute Deviation (MAD) (~ a sort of

deviation) of a data set is the average of
the absolute deviations based on a number of
observations from a central point. It is a summary
statistic of statistical dispersion or variability.

48
 o We get rid of the minus and plus, by applying the absolute value.
o The mean absolute deviation, the smaller it is, the better it is.
• Mean Squared Error (MSE) (~ a sort of a variance): measures the average of the squares of the errors—that
is, the average squared difference between the estimated values and the actual value. MSE is a risk function,
corresponding to the expected value of the squared error loss. The fact that MSE is almost always strictly
positive (and not zero) is because of randomness or because the estimator does not account for
information that could produce a more accurate estimate. The MSE is a measure of the quality of an
estimator—it is always non-negative, and values closer to zero are better.
• Mean Absolute Percent Error (MAPE): it is a measure of prediction accuracy of a forecasting method
in statistics, for example in trend estimation, also used as a loss function for regression problems in machine
learning.

Example 4:
1) Calculate actual - forecast for each value of both
models, therefore MAD for each observation and sum
them up (|(52-55.0)+(52-54.7)…|)
2) Divide the total by 10.
3) By looking at the results, it’s clear that Model 1 is
better since it’s lower, so closer to the actual value.
Moreover, it’s better to consider the lowest value in
case you’re using it for safety stock level, it would
generate lower costs.

Which
department in a Company performs forecast? Marketing and sales – this is why it is called the “demand side”.
Forecast is also defined as planning. The independent demand is forecasted, then the company hands this over to the
production people, so the independent demand becomes now to a dependent demand, because this is the number
production people are using all the decisions and dependent demand means that now we can control this: if this is the
number, we can decide whether to spread it over the production period, or to produce it when the order comes in.

(we’re in the operation side of the company)

AGGREGATE PRODUCTION PLANNING

It’s a marketing activity that does an aggregate plan for the production process, to give an idea to management as to
what quantity of materials and other resources to be procured and when, so that the total costs of operations\ of the
organization is kept to the minimum over that period.
The goal of an aggregate production plan is to specify the production rates, inventory, employment levels, backlogs,
possible subcontracting and other resources that are needed to meet the sales plan. The overall goal is to minimize
the involved costs:
• Inventory holding cost
• Regular production cost – average labor costs to produce an aggregate unit
• Overtime cost
• Hiring cost
• Firing/layoff cost

49
• Backorder/lost sales cost
• Subcontracting cost

Aggregate demand exercise

240,000 cases of soft drinks, the period covers 6 months,
and we have to make sure that these numbers are
available when they are required. The planning values
report the information required to make the aggregate
demand.

Alternative production plans

In this figure, there is a level production plan, and in the first

period plan until the end of March the level of demand is lower
than the production level, so we accumulate stock.
In April, May and June, the demand is higher than the production,
so we’ll use the stock accumulated in the previous period to satisfy
the demand.
What costs have to be considered for the level production plan?
Holding costs, since we accumulate stock in the first period and
production costs.
What other costs could be incurred?
Firing and hiring costs. Hiring costs occur when the workforce has to be increased and training is included. Firing costs
occur when the workforce has to be reduced.

In this figure, there is a chase production plan, production

follows demand – the company always adapts resources to
demanded quantity in the single period.
Which costs do not occur in the chase production plan?
Holding costs.

Translate the sales forecast into resource requirements

What are the regular labour costs per case?

160ℎ × 20$
= 0.80$ 𝑝𝑒𝑟 𝑐𝑎𝑠𝑒
4000 𝑐𝑎𝑠𝑒𝑠
What are the overtime wage rate costs per case?
0.80 × 1.5 = $1.20 𝑝𝑒𝑟 𝑐𝑎𝑠𝑒

50
 Level production plan
There is a stable production, 40,000 cases per
month, that is 16,000 more than required for the
first period.
For the second period, production is 8,000 more
than sales forecast.
In April, however, sales requested are more than
production, so I can use the inventory held on
stock.
Costs to be considered are production, hiring
and holding costs.

Figure: Costs of level production plan

Chase production plan

Workers are changing overtime. In this case here costs
incurred are hiring and firing costs, production costs and
holding costs.

Figure: Costs of chase production plan

Mixed/hybrid production plan

Inventory is used to satisfy as much as possible

and everything above is either overtime or
subcontract. We produce internally 192,000, in
this case we also have extra inventory. In April
there is a 16,000 of difference, and everything
I have to produce more is done by
subcontractors.

51
Comparison of plans

INVENTORY MANAGEMENT
Inventory management is important for every other business enterprise.
• Economic order quantity (EOQ) – single product/single price
• Economic order quantity (EOQ) – single product/quantity discount
• Economic order quantity (EOQ) – multiple products/single
price – so in an order we have multiple products, and we treat
these products individually.

Inventory management focuses on how much to order and the

delivery time. Through the graph we can see how the inventory level
changes, and the reorder point is the quantity required to satisfy the
demand that occurs during the delivery time. After the reorder point
there is an increase the order quantity, and this is called economic
order quantity or lot size x.
This is the general idea of inventory management.

What is inventory/stock?
Those stocks or items used to support production (raw materials and work-in-process items) supporting activities
(maintenance, repair, and operating supplies) and customer service (finished goods and spare parts). Inventory is a
logistic driver.

The focus, with inventory management, is in quantity, and what is the cost and value beyond it. Even though, we use
inventory and stock are used as synonym, inventory is measured in terms of value and stock is measured in units.

52
Inventory for deterministic (constant) and stochastic demand

Example

This quantity is very constant, so we have a linear decrease,

and whenever we turn to zero, we receive a quantity and
there in an increase, and so on. On average we have half of
the quantity Q (Q/2) always on stock: the so-called cycle
stock. With such a pattern, no matter the quantity ordered,
half of the quantity will be held in the inventory and for this
quantity, inventory holding costs have to be calculated.

Which costs are related to inventory?

• Product cost
o The amount paid to suppliers for products that are purchased
o Variable cost since it depends on the quantity bought from the supplier
• Carrying/Holding cost
o Expenses that are incurred due to the fact that inventory is held
o Variable cost since it depends on the amount of inventory held
• Order and setup cost
o The expenses incurred in placing and receiving orders from suppliers as well as administrative
expenses to set up an order
o Fixed cost since it doesn’t depend on the quantity purchased
• Stockout/shortage cost
o Cost incurred when inventory is not available to meet demand
o Variable since it depends on the quantity

53
How to find the quantity that lower these costs?

On the x-axis there is lot size and in the y-axis costs. The
material costs are not changing, holding costs are high when
lot size is high and ordering costs; I have a small lot size,
which means I’m ordering a lot, so high ordering cost. If
instead, I order a high lot size, I’m ordering just a few times so
ordering cost is quite low. The curve is the sum of these three
costs positions, and at a certain point, there is the minimum
where holding costs and ordering costs are crossed.

Total annual cost calculation

Solution Inventory Management 1a+b

If the first part is not changing, then why we calculate this? So, we get rid
of that, and we only look at ordering costs and inventory holding costs.
Through this, we notice that I order more but I have less holding costs
and vice versa.
This leads to that we only to fixed order costs and inventory holding
costs. The curve is not so steep. We’re only looking at the so-called
decision-relevant costs; we want to minimize the sum of ordering costs
and holding costs.

54
In order to minimize costs, I have to order 273.86 units.

Now what happens if purchasing prices change?

Quantity discounted model

I have to check whether the quantity a is within the 0 -1,000
range but what if quantity a doesn’t stick in this range – then I
cannot use that quantity with that price, because I couldn’t reach
that price level. In order to reach it, it’s necessary to increase
order quantity to the minimum level of the price range, and that
would be my order quantity – I will get 19.00$ with a minimum of
1,000, no less.

In order to get 19, I have to increase up to 1,000.

Determinations of total costs – how are they different as compared to the first EOQ model?
This means that all costs have to be included and we always have to choose the lowest level of total costs.

55
What is the replenishment strategy with the lowest annual cost?
• Ordering independently from each supplier or aggregated for joint delivery?
• What would be the cycle inventory of each component at Harley?
What if multiple products are aggregated in a single order?
• Savings in transportation costs
o Reduces fixed cost for each product
o Lot size for each product can be reduced
o Cycle inventory is reduced
• Single delivery from multiple suppliers or single truck delivering to multiple retailers
• Receiving and loading costs reduced

Lot sizing with multiple products or customers

In general, ordering, transportation and receiving cost of an order grows with the variety (number) of products or
pickup points. Lot sizes and ordering policy that minimize total cost
• Di:Annual demand for product i
• S: Order cost incurred each time an order is placed. independent of the variety of products in the order
• si: Additional order cost incurred if product i is included in the order.

What are the costs for individual EOQ?

OQ Ordered and Delivered Jointly

n = number of orders that I made

56
Aggregation of Harley Division

Aggregation in this case helps to lower costs.

lesson 11

26.02.2021

Warm up exercise

Ø EOQ answers how much – so the quantity that covers the demand
during the lead time
Ø ROP (reorders point) tells when to order
7 :;<
Ø Total cost = :;< × 𝑜𝑟𝑑𝑒𝑟𝑖𝑛𝑔 𝑐𝑜𝑠𝑡 + 2
× ℎ𝑜𝑙𝑑𝑖𝑛𝑔 𝑐𝑜𝑠𝑡 𝑟𝑎𝑡𝑒

57
Example exercise

Solution: The EOQ has to be calculated for each order range. The first one has been excluded since the order quantity
didn’t fit between 1 and 49 (126.49), so we excluded then for the second we obtained 133.33 (√ (2•600•20) /
(0.3•4,5) = 133,33) and it fit between 50 to 249, so the holding costs, set up costs, unit costs and total costs have been
calculated.
Why does the first range not fit? Because the EOQ is higher than 49, so if you order 126.49, you’re not going to spend
5.00$ per unit but 4.50$, because it would fall in the second range à this is why it doesn’t fit, and of course you
cannot lower quantity within the range.
In the third range, it doesn’t fit either, therefore the second one has been chosen. So, with 139.69 as a quantity, the
unit price is 4.50$; in order to obtain 4.10$ as a price, at least 250 quantity units have to be purchased. Therefore, by
considering 250, it is possible to calculate the other costs. Since it’s a big quantity, there will be higher holding costs.
In this situation here, I always want to get the best price, so I have to raise quantity up to 250 units, in order to have
lower total costs. This doesn’t mean that the low price gives the best total costs situation: it’s fundamental always to
check out where are the lower total costs, by calculating them.
à I’m always looking for the best price that gives me the lowest TC.

INVENTORY MANAGEMENT – SAFETY STOCK

- News-Vendor Problem
- Periodic review system
- Continuous review system

The basic idea behind safety inventory

• Safety inventory helps to improve product availability in the presence of supply and demand variability:
helps us improve the availability of my product.
• Supply chain managers are aiming at maintaining/improving product availability by reducing the amount of
safety inventory: they want to do this with a lower safety stock level. Information helps to reduce inventory
and we also learned that one issue….
Typically, a supply chain consists of a lot of stages. The first example of bullwhip effect is Barilla in 1990. Barilla is
producing pasta and the retail environment is really special in Italy, so barilla as a manufacturer is selling the product
to wholesalers and wholesalers sell to distributors and distributors sell pasta to small shop.
If you have a lot of levers in between, you often don’t know what is going on the retailer level. You don’t know how
the demand is developing.
The more actors are involved the more they do not interact the more they send orders that do not reflect the correct
market development, the more the information that goes upstream is distorted so you have amplifying developments
so, this is called the bullwhip effect.

58
The demand in Italy for pasta is stable but you want to get a better price so that’s why you order more.
Super important issue for supply chain management: I can improve all these things if I exchange information between
actors and I harmonize my activities and I have access to the retail level or to the consumer level and see how demand
is really developing.

Safety inventory is carried to satisfy demand that exceeds the amount forecasted – the so-called unexpected
demand:
- Raising the level of safety inventory increases product availability and thus the margin captured from
customer purchases (this is the good side).
- Raising the level of safety inventory increases inventory holding costs (this is the bad side).

Three key questions managers ask themselves:

1) What is the appropriate level of product availability? We’ll see in the situation of a single period model, we
compare the changes of having too much against the changes of having too less, to have the so-called profit
critical ratio, which is the eservice level at which I get the highest profitability.
2) How much safety inventory is needed for the desired level of product availability? if I’m setting a customer
service level which is called the desired level of product availability, in a business world you agree with your
client, if you send me 100 orders, I guarantee you that 80,85,95,99 out of 100 will be executed exactly as you
want, and you accept a failure rate of 1% of order that is not fulfilled as you want. Once you this agreement
which is called service level agreement then you can identify how much safety stock, I need for pleasing
90,95,99 costumers out of 100.
3) What actions can be taken to improve product availability while reducing safety inventory? Then I will see
that sometimes I’m able to improve my product availability even though I can have lower safety stock level,
and this involves cooperation with supply chain partners. Where in the supply chain can I have safety stock? I
can have it in other places too, maybe it’s not me who needs to have safety stock but it can be somebody
else.

Here we have an extra layer, compared to the previous

graph, that is safety inventory level, which allows to have
available stock for necessity – we’re still able to satisfy
customers demand if orders do not arrive.

How can we determine the appropriate level of safety

inventory?

The appropriate level of safety inventory is determined by two factors:

1) The uncertainty of both demand and supply

2) The desired level of product availability

Measuring Demand Uncertainty

D = Average demand per period
SD = Standard deviation of demand (forecast error) per period

Lead time (L) is the gap between when an order is placed and when it is received

What causes variability?

• The variability of the demand
• The variability of the lead time
• The average length of lead time
• The desired service level.

59
Normally distributed demand curve

The dot line in the graph that divides the curve represents the mean of the
distribution into two halves.

If you order the mean demand, how much of the total demand do we cover?
50%. What is the safety stock? It is 0.
If instead, I want to cover more of the 50% of demand: how many standard
deviations away from the mean do I want to cover? Therefore, the safety
stock tells me the number of standard deviations I want to have in order to
cover a certain share of total demand.
Here, in this case, we want to cover 95%.
If I want to cover 68% this is one standard deviation. If I have one standard deviation, I’m able to improve my product
availability by 20%.
If I have 2 standard deviations of safety stock, I’m about 99%. If I want to cover 99,9% I need 3 standard deviations of
safety stock.
What does it mean? The more I go into this area (blue one), the increases of improvements in terms of getting more
clients, any additional percentage of sales volume will lead to significant higher increases in building up safety stock,
because I have to add more and amore standard deviations of my mean demand.

(2/03) We’re looking at a stochastic demand and the other demand we want to look at has deviations. When there is a
deviation and it’s fluctuating, we cannot be sure that the number of customers show up daily is not really precise;
what I want to do is to find the share of the total demand that gives us the highest probability. We calculate the area
that we would like to cover. In other cases, we make a type of decision, we want to guarantee at least a certain share
of our demand that our clients receive the quantity desired.
So, we have to decide the area to cover and the risk of stockout. Then, we have to decide the number of standard
deviations. Subsequently, we look at the tables.
I always take at the lowest one, in the z-table, between the two, which is equal for example to
=0.95. In this case, we’re looking for 0.95 in the table, and between 0.94950 and 0.95053, choose
always the lowest one. So, it is 1.64 standard deviation away, based on the lowest number in the table, closer to 0.95,
so 0.94950.

Safety stock = average demand that we observed + Z* standard deviation.

Calculating a service level

Service Level: A term used to indicate the amount of demand to be met under
conditions of demand and supply uncertainty.
Assume that the demand during the reorder period and the order lead time is
normally distributed.
Table on the right: if I want to cover 90%, this will cost me 1.28 standard deviation of
safety stock.

Example: How many newspapers am I going to sell today?

If we do not know how much to sell, how do we know how much to order? It is sufficient to know with which
probability different demand levels occur!

Example: How much should we order per day? Use the News-Vendor-Approach

Suppose you open a kiosk at the mall every October to sell Halloween costumes. The most popular costume
historically has been a skeleton costume:

60
 • You can buy the costume for $10 and sell it for $30.
• Any costumes not sold have to be disposed of because the design changes each year and customers will not
purchase a previous year’s costume.
• Disposal and salvage costs are minimal and can be considered zero.

So, what do I have to think?

What are the costs of a non-sold item that cannot be sold tomorrow? à Costs of
overstocking
What are to costs of an item I cannot sell because I do not have any more of this
item? à Costs of understocking

News-vendor replenishment quantity (single period system)

Figure: Solution
Determining the order quantity

If 100 customers come, 67 will obtain the custom, if 200 come, 67% will come. If I buy 206 customes, if I order 220, I
have a lower profitability and if I order less than 206 I have lower profitability. With 206, I have the highest
profitability.

lesson 12

2.03.2021

Managing independent demand inventory – control systems

Periodic Review System =PRS
- The inventory level for an item is checked at regular intervals and restocked to some predetermined level
Continuous Review System = QRS
- The inventory level for an item is constantly monitored and when the reorder point is reached an order is
released.

61
What is the reorder point? It’s the demand during the lead time, and since the demand is fluctuating (not constant),
we have to cover uncertainty due to fluctuations. In PRS, we regularly checked with fixed intervals, how my inventory
looks like and to know it, I have to start by identifying a predetermined level, seen as a maximum level and known as
restocking level. In case I control every 7 days in the supermarket shelves, how many boxes of pasta I have sold, I need
to know where we started the last time I’ve been there. Once I came up with a level, that allows me to cover all the
demand that will be in those 7 days and I have to include the demand that is going to be while I’m waiting for the
order. In this time span, the ideas of restocking point and safety stock are included. How does it work? Look at the
example below.

In this case we want to know the safety stock for 9

days, given the stock 99% and we want to know how
much we’re going to sell in those 9 days.

Therefore, the restocking level (R) in a periodic

review system:

2.33 is the safety factor, in the normal distribution table, obtained by

checking in the table 0.99.

We identified 155, the maximum value, if we have 100, we have to fill

up with other 55.

The restock level and the current inventory level are compared. So, if our inventory level is currently 45, in order to fill
up the inventory (which capacity is 155), we need 110 more units.

How is the reorder point calculated

The first one occurs when demand is stable, while if demand is fluctuating SS has to be added (standard deviation =
safety stock).

62
Example Continuous Review System

1) The daily demand of a product is 10 units. Your supplier delivers your order 9 days after an order has been
placed. At which minimum inventory level is it necessary to trigger an order?
• What is the reorder point? 90 = d*L= 9*10

2) The average demand during the lead time of a product is 90 units with a
standard deviation of 25.4 units. The agreed service level is 95%.

Inventory Management 3

A local club is selling Christmas trees and deciding how many to stock for the month of December. If demand is
normally distributed with a mean of 100 and standard deviation of 20, trees have no salvage value at the end of the
month, trees cost $20, and trees sell for $50,

• What is the service level?

a) .60
b) .20
c) .84
d) .40

This is a news vendor problem: overstocking or understocking could occur, therefore I have costs of overstocking
(costs of having too much) and costs of understocking (costs of having too less). The costs of overstocking are 20-0,
while the understocking costs are 50-20, where 20 is the purchasing price, 0 is the return price (salvage value) and 50
is the selling price. So overstocking is 20, understocking is 30. How do I get from this to the service level? I try to
balance the risk of having too much and the risk of having to less à 30/(30+20)=0.60. 60% is the highest profitability.

Inventory Management 4

A local barbecue joint makes one massive batch of potato salad each day. If they run out of this savory side before the
end of the day, their last few customers are less than satisfied, but if they make too much, they sell it to the local hog
farmer for feed. Every serving costs an equivalent of $0.23 to make but can be sold for $1.50 to customers and for
$0.08 to the hog farmer. The average daily demand is for 200 servings with a standard deviation of 20 servings.

• What is critical ratio?

• How many servings should be made each day?

Purchasing cost is $ 0.23, the return price is $ 0.08. 0.23-0.08=0.15, costs

of overstocking. Sales price is $1.50 and purchasing price is 0.23$,
therefore the costs of understocking is 1.27$ (1.50-0.23). Therefore, the
critical ratio is: 1.27/(1.27+0.15)=0.89. Now we look at the table, and we
check z (1.22), to identify the safety stock. S*= u+z*standard deviation.

Inventory Management 5

The Winfield Distributing Company has maintained an 80% service level policy for inventory of string trimmers. Mean
demand during the reorder period is 170 trimmers, and the standard deviation is 60 trimmers. The annual cost of
carrying one trimmer in inventory is $6. The area salespeople have recently told Winfield's management that they
could expect a $400 improvement in profit (based on current figures of cost per trimmer) if the service level were
increased to 99%.

63
• Is it worthwhile for Winfield to make this change?

mean demand = 170, standard deviation =60,

S = 170 + 2.33*60 = 309.8 (99%)

S = 170 + 0.85*60= 221 (80%)

It’s better to remain on the 80%, since we should have more

safety stock in 99%, and therefore it’s more expensive. 532.8 $
would be the extra cost of the additional safety stock.

NETWORK DESIGN (FACILITIES)

Facilities are the places where we produce or store our products.

Where to locate the facility? There are multiple factors to consider:
• Labor: do we have access to different labors?
• Proximity to suppliers and customers
• Cost of land and construction
• Taxes, incentives, and regulations
• Transportation infrastructure
• Quality of life for employees
• Supply chain risk

If you had to open up a central distribution center in Europe to deliver to all European countries, in which country
would you place this facility? A country with a port.

Why has Red Bull two production facilities in such a near environment (one in Austria and Switzerland)? European
products wouldn’t be able to enter the US market, while if products come from Switzerland, they can overcome
barriers in US and enter there.

Response Time = how much does it take to respond to an order. Of

course, if the number of facilities is high, the response time to demand
of customers is lower, even though costs are higher (holding costs,
inventory etc.). This is why, it’s better to concentrate facilities in a
country, to avoid high costs.

If you’re in a situation in which you have to

identify a specific location for a distribution
center, once you’ve identified a country and
identified customers, the general description of
the problem should be the following.

64
Try to solve this problem: A county wants to build one centrally located processing facility to serve the county's three
recycling drop-off locations. The three drop-offs have characteristics as given in the table below.

With the Center-of-Gravity method is possible to find the best

location.
With this method we need to know:
- Location of markets
- The volume of goods shipped to those markets
(customers)
- Shipping cost (or distance) – higher the distance, higher the effort, of course.
The purpose is of course to find a cost-optimal location for a facility.

Center-of-Gravity method: finds the lowest-cost location based on

demand and distance, using X and Y coordinates to define a geographic
position.

Therefore, by looking at the

result, do we set up the
facility exactly at 54.5 on the
x and at 50.6 on the y? We
have to check whether there
is an appropriate land available in that location (maybe there is a lake, sea
etc.). This is the disadvantage of the center-of-gravity model, while the
advantage is that is really immediate and easy to compute. Therefore, the next step is to check whether is really
possible to set up a facility there; if not, choose a place close to it.

TRANSPORTATION PROBLEM

Once we’ve set up the facility, we look at transportation.

What is the problem?

Which facility is going to deliver certain quantity to a certain level of customers? (cost-minimal connection) -I’m
looking to the allocation of products produced in facilities to customers’ demand.
We have the following information:
- Location of plants and their capacities/quantities of a given period
- Customer locations and they required quantities
- Delivery costs between delivery and receiving points (so transportation costs).
What would be the least costly allocation of facilities’ products to customers’ demand?

Example: We’ll start with facilities (4

sources), customers (4 destination) and
the production (supply). Source 1
produces 105 units, Source 2 145 and
source 3 50. Destination 1 demands 80
etc. We cannot ship more than 300 units
and customers will not ask more than 300
units in total. The numbers in the circle are the shipping costs/transportation costs to ship units from a source to a
destination. Therefore, it costs 12 $ to ship 1 unit from source 1 to destination 1, 18$ to ship 1 units etc.

65
Question: how to allocate supply to demand by minimizing costs?
Method to be applied: heuristics is a mathematical approach that gives a solid solution.
Possible solutions:

Initial solution through Northwest-corner rule

Start in the upper left-hand cell (or northwest corner) of the table and allocate units to shipping routes as follows:

• Exhaust the supply (factory capacity) of each row before moving down to the next row
• Exhaust the (warehouse) requirements of each column before moving to the next column
• Check to ensure that all supplies and demands are met

We want to allocate the supply to the demand. We start with

80, we give it to them, the second person wants 25, and I still
have them, so I give it to him (105 completed).
Then I go to the second source, the customer 2 requires 35
units, I give them to him, the third customer 70, I still have
them and lastly 40, which I have them (145 completed). And so
on… Z represents the costs.

How to come up with even a better solution? àSteppingstone

method

1) Select any unused square to evaluate

2) Beginning at this square. trace a closed path back to the original square via squares that are currently being
used.
3) Beginning with a plus (+) sign at the unused corner. place alternate minus and plus signs at each corner of the
path just traced.
4) Calculate an improvement index by first adding the unit cost figures found in each square containing a plus
sign and subtracting the unit costs in each square containing a minus sign.
5) Repeat steps 1 though 4 until you have calculated an improvement index for all unused squares. If all indices
are ≥ 0. you have reached an optimal solution.

Steppingstone method example:

The blue numbers came up with Northwest

corner rule. Once I completed this, I can
reason on improvements, and I can suppose
and think that empty cells may be better: how
do I find out? By comparing the costs.
Therefore, we basically check out (this is the
steppingstone method) by taking for example the squared section, how do we know if Source 2-Destination 1
wouldn’t be a good idea respect to delivering, as it happened, from Source 2 to Destination 2? The method says, I
check out if by moving one piece that is moving out to the next destination in the empty cell, gives an improvement or
not – if it gives an improvement for one unit, it will generate improvements for all units. Let’s consider costs and
compare them, to determine whether it generates benefits or not; however, some restrictions have to be considered
– we represent them through the step 3, concerning putting plus and minus in alternate ways. Then, I can determine,
what are the shipping costs in the cell (add shipping costs in cells with plus, subtract them in cells with minus)? If the
total result turns into a negative, we have safe costs, if it comes to a positive result, it is adding costs.

66
These tables plot what we’ve just found. The second table and result mean: If I ship from Source 2 to Destination 1 my
cost will go up: one unit will cost 20 € more. If I do this
procedure for all the empty cells, I can see net cost changes.
From S1 to D3, I have a saving of 32€. I do this procedure
again and again, until I meet positive results: once I receive
only positive numbers, the previous one was my best solution
and then I stop.

Therefore, the best solution is:

VEHICLE ROUTING PROBLEM

We have now the orders, the question now is: in which sequence we’re going to meet our clients? Of course, you
want to minimize costs and typically you start from a starting point, you meet all clients and go back to the starting
point. The aim is to minimize time, costs and effort.

How to allocate vehicles to delivery routes?

• I vehicles are used to deliver a number of customers from one warehouse.

• Which vehicle shall be used for which customer and in which sequence shall vehicles go to customers?

This is what Amazon responds every day.

Example

• 5 customers with demand of one pallet each

• Trucks with max 4 pallets each

In L6 we have the warehouse, and we have 5 clients to visit (A1,A2,A3,A4 and A5)
and the restriction is that on trucks we can carry a maximum of 4 pallets.

67
1)How many vehicles I need? 2)And in which sequence should the clients be visited? The first approach to find a
solution, is the initial solution: I go to the first client, then I return, then I go to the second, then I return etc.
Therefore, you have single deliveries. Costs are of course high. The second solution would be
to go to the first client, instead of returning I go to the second client and then I return, and so
on. If I follow this procedure, I’m able to reduce some returning costs and delivering costs
(here, I avoid the returning cost from A1 and to deliver to A2). So consider L6-A1-A2-L6: I have
a saving of -30-35+15= -50, so the saving is -50€.
❑Initial solution: Z0 = 2 * (30 + 35 + 40 + 30 + 60) = 390

Figure: Cost changes

The rule àThe savings heuristics

Network Design 1

A company has six customers and wants to set up a distribution center out of which the demand for those customers
will be satisfied.

What is the best location for the distribution center? Use the Center-of-Gravity

68
Network Design 3

A brewery has to deliver its products to six customer warehouses. The brewery can use a total of three trucks with a
loading capacity of max. 20 pallets each. Following table shows the distances between the brewery and the
warehouses. The number in brackets indicate the demand of the individual customer in pallets.

What are the costs for the initial solution? And what is the savings for consolidating customers 1 and 2?

lesson 13
5.03.2021
Warm up exercise – center of gravity
- A company has six customers and wants to set up a distribution center out of which the demand for those
customers will be satisfied.
- What is the best location for the distribution center? Use the center-of-gravity.

Center of gravity looks at all the location models. Any facilities model is looking for a point in the space, and out of this
point, the delivery to all the customers and satisfaction of their demand is guaranteed to be at the lowest level of
effort. Therefore, if we deliver all customers’ demand, then the total effort to satisfy these customers is at the lowest
level.
With the center of gravity approach, it’s necessary to know just where customers are located and what is their
demand. When we consider the distance between the potential distribution centers and the customers, the longer
the distance, the higher the effort and vice versa. In this specific case, moreover, the higher is the demand, the higher
is the effort. Therefore, need to find a point which can balance this out.
How do it? We should multiply the location for each Customer X to the customer Demand, divided by the customer
demand and then do the same thing for Y and choose the lowest one.

5 × 31 + 8 × 28 + 6 × 53 + 14 × 32 + 10 × 41
𝑥= = 7.62
31 + 28 + 19 + 53 + 32 + 41

13 × 31 + 18 × 28 + 3 × 53 + 20 × 32 + 12 × 41
𝑦= = 10.77
31 + 28 + 19 + 53 + 32 + 41

69
These are the coordinates of where we’re going to establish our facility; however, we have to see if there are any
external constraints and check availability à (Is this a good area, is there place to locate the facility etc.?)

Northwest corner rule exercise (papabile)

For the problem data set below, what is the northwest corner allocation to the cell Source 1-Destination 1?

A) 0
B) 2
C) 15
D) 90
E) 30
Destinations = customers – with a certain demand (they don’t desire more than that quantity).
Sources = factories – with a certain supply (we don’t ship more than this quantity).

The numbers are transportation costs.

We want to allocate our total quantity of 90 to 90 demand, by reducing costs at a minimum level, so that
transportation costs are the lowest value.
The northwest corner rule is used to allocate, while the minimization of costs is reached through steppingstone.
Allocation:

The answer is C.

Steppingstone exercise - Network Design 2

Consider the transportation problem and its initial solution in the table below. What is the improvement index for the
empty cell Source 1 – Destination 3?

All the empty cells represent improvements of the optimization.

70
With the improvements we compare costs.

-14 is good since it reduces the costs, and you’ll be saving money.
What is the next step? We need to continue use the steppingstone until all the new solution will become positive.

Solution:

Vehicle routing problem exercise

- 6 customers
- 3 trucks with maximum 20 pallets each

The numbers in the table represent costs: for example, we would have a cost of 17 to reach client 1 and again 17 to
come back.
The goal of the vehicle routing problem is to allocate vehicles to customers, to find out how many vehicles do we need
to serve customers. Once the allocation is conducted, the order should be set (the order to visit customers).
To solve this problem, we need to find a starting point and then look for savings.
How would the initial solution in this situation look like?
We could go from WH to client 1 and come back to WH, then go to client 2 and come back: (17+25+22+27+17+29)*2
=274
How do I come to improvements?
Instead of going to client 1 and return, and then go to client 2 and return etc. what could I do?
Go to client 1 and client 2 and then return.
= -17-25+10= -32
Therefore, I have a savings of 32.
Solution:

71
 lesson 14

9.03.2021

PHYSICAL DISTRIBUTION

The physical distribution of P&G, which produces detergents for national

market, takes place in the last part of the supply chain, so on the
downstream part: once the finished product is ready it is delivered to a DC
(distribution center) which is a sort of warehouse that covers a geographical
area, and later one it is moved to one of the stores of Walmart and in the
final stage is when the customer purchases the detergent in the store.

Design of the physical distribution system

There are different choices that have to be taken when you design the physical distribution system.
Before considering the choices, let’s consider the physical distribution system of two companies belonging to two
different industries: luxury companies’ most common physical distribution system involves that the company has just
one warehouse in Europe and serves all the European stores. The second example is the one of Coca-Cola: the
production of this brand is conducted in a few plants, mostly located in US and this kind of syrup is put into bags, sent
to the National Bottling Company (that put water in the syrup to dilute it and then, take care of the bottling process) –
this is done to protect the recipe. In Italy, there is just one bottling company, in Veneto; but how are clients reached
(supermarket, bars, restaurants etc.)? Once the bottling process is completed, the product will be stocked in the
factory close to it and then moved to many distribution centers, that cover different regional areas (one for Lazio, one
for Lombardia etc.), then another delivery is conducted to reach the final customer. This means that the Coca Cola’s
physical distribution system, has many logistics centers, next to the factories, and many distribution centers that cover
specific regional areas. If you compare Coca Cola and fashion industries, in the latter there is one warehouse for the
entire house, while for Cola-Cola there are many distribution centers for the national country: so, is there a problem?
Both systems can be appropriate for their industry, so when you design the physical distribution system, there is not
unique solution that fits all situations, but variables have to be considered, to choose the most appropriate one.

Stages of the pyramid to design of the physical distribution

system:
1) Strategic Choices: concern the customer service
(service that we want to deliver to the clients) that the
company wants to deliver. It is defined in terms of Total
Logistic Cost and of Service Level
a. Total Logistic Cost: Transportation, Holding
costs, Warehouse costs, Administrative costs.
b. Service Level: Speed, Dependability, Availability
and Completeness and Flexibility
Of course, when we choose the customer service, we have to
decide on which points to focus more (speed, dependability etc.)
by considering also costs. So of course, there must be a trade-off
between total logistic costs and service level, because there are cases
in which we cannot afford all the elements we desire.
In physical distribution, some strategic choices can be implemented:
in this matrix, we have service level and total logistic costs, with high
and low level, and based on their level we have different strategies.
The only combination that can’t be considered a strategy is low

72
service level and high logistic costs: you are not fast in delivery, you are not good in dependability etc. and you also
spend a lot. The others are instead strategies that can be chosen.
- Marketing Oriented strategy (high service level, high logistic costs): we need to charge higher prices to
clients, in order to receive more complete deliveries, faster deliveries etc. Example: home delivery, when you
buy grocery products online. Besides the cost of the order you made, you also pay for the shipping, in order
to have it faster, more dependable etc. therefore, in order to receive this service, you spend more money.
Another company is Caterpillar that produces machines like scrapers, which is able to deliver all over the
world very quickly: this is a good service for clients, because by receiving products in a short amount of time,
they don’t have to stop the production for their companies. In order to provide this service, of course more
money is required, so this service is quite expensive (stock available all over the world, costly modes of
transportation etc.). However, this is a situation in which the client is still willing to pay, since the payoff is
very high.
- Production Oriented strategy (low service level, low logistic costs): the supplier won’t be fast, won’t focus on
dependability etc., but at the same time, the customer will spend less. Example: Suppose that a client wants
to buy a bookshelf in Ikea, but he doesn’t want to spend a lot of money. The customer in the Ikea’s
warehouse takes all the components, and in order to save money, he takes care of and transportation and
assembly. By doing this, many logistic activities, like picking process, transportation and final assembly are
outsourced to the customer (so they are for free), therefore Ikea reduces the level of cost.
- Logistic Oriented strategy (high service level, low logistic costs) best solution: the customer will be very
happy, since the service level is high, and he doesn’t have to spend so much money which is coupled with
high service level in logistics. This is the clear example of Amazon, since it offers great services at an
affordable price. The great advantages of Amazon are fast delivery, high level of availability, prices are
actually quite low. How it can do that? They sell online, so Amazon doesn’t have stores and linked costs, just
a few warehouses and distribution centers. The warehouses are also in the countryside, so also the costs in
purchasing those areas are very low. Moreover, when it organizes any process related to logistics, Amazon is
very clear, since it uses the “organized chaos” to place products in the warehouse: this concept means that
Amazon places all products in the warehouse without dividing them for categories, because if I put all the
shampoos together, when I receive an order is easier to make a mistake if they are all put close to each
other. This is why Amazon puts products in a chaotic way. In order to reach the logistic oriented strategy,
multiple tentative have to be conducted.

In the perspective of the client, the most relevant performances of

strategic level are speed and dependability, so companies strive to find
better time performances. There can be cases in which companies
decide to focus more on dependability respect to speed (time definite
systems), or more on speed respect to dependability (time critical
systems), like Amazon. Time Based System occur when both
dependability and speed are important, this is the most challenging
situation.
2) Structural Choices: which are generally two:
- Channel design: as we know, channel distributions, generally speaking, can be direct or indirect, where the
direct one occurs between producers and clients when there isn’t any other player (ex. when you buy a Gucci
bag, from a Gucci shop), while the indirect occurs when there are some intermediaries between producers
and clients (ex. you buy Barilla in the supermarket chain, which is the intermediary between the Barilla’s
brand and the clients). There can be just one intermediary, so just one retailer, it is an indirect short indirect
distribution channel, while if there are several intermediaries, it is called long indirect distribution channel
(when for example, Barilla is sold in another country).
Generally, the decision of distribution channel is conducted by the marketing department.
The online channel is a new channel of distribution, which is quite complex because of the “last mile”
logistics, which refers to the home delivery of the products bought on-line, that is complex and costly as well.
Home delivery is complex because products have to be delivered to many different places, so there are

73
 multiple customers and different address (they can live in streets, mountains, countryside etc.). On the
contrary, if you sell through a traditional channel distribution, it would be a duty of the customer to come to
the store. The problem of the last mile logistics was initially overlooked. On the other hand, there are some
new solutions that have been introduced to lower the complexity of online channels. This is the case of the
click and collect service, which is a very easy system, that allows to reserve a product for a customer, that will
pick it up directly from the store – this is good also for the customer if the store is close to his house and he
avoids home delivery that requires waiting and also to be at home. Another solution is Amazon lockers, with
just one delivery the courier places dozens of orders and can be reached in any hours, while home delivery
cannot be conducted at 6.00 a.m.
- Network design: The logistic network is the bundle of warehouses and distribution centers (DC) that the
product passes through before reaching the customer. The main
network design choices are the following:
o Centralization/decentralization: this is a decision about
the number of warehouses and DC needed in the
system. How to take this decision? In the photo on the
right, there are two alternatives: one with high value
density (high centralization, few numbers of
warehouses and DC – Gucci, Ferragamo etc) and the
other with low value density (low centralization, one warehouse and products before being
reaching the customers, are sent to different DC placed in different geographical areas – Coca-Cola
case). Centralized system (first figure): there is more control on products (quality etc.), even though
a huge number of responsibilities are condensed. Let’s consider you have to arrange the delivery of
some luxury bags from Milan (where the WH is) to London (where the shop is); of course, the
delivery wouldn’t be too large, since in London there is just one shop. By planning the delivery with
the courier, I can afford the transportation, even if it may be a little bit high, because the price at
which I sell the bag is very high and the weight is relatively low, so the turnover would be quite high.
If instead of having luxury bags, I have Coca-Cola bottles, the value of a box of six bottles is 12€: does
it make sense to spend 26 € in transportation? Of course, not. This is the reason why, Coca-Cola uses
a decentralized system, since the level of efficiency in transportation is much higher, since in the
long routes between WH and DC, I move a large number of quantities for a regional area in which DC
is located; I’m no longer sending just a few items for a shop.
Decentralized system: there is a movement of large quantities from WH to DC, so transportation
costs are lower (it’s a sort of economies of scale). Then a small delivery has to be conducted from DC
to final customers, but the delivery covers just a small area, since it’s the area close to the
distribution center.
Value density is a ratio unit value/weight or the volume: fashion handbags have a high unit values
and weight extremely low. If the price is high, but weight is low, transportation costs may be
irrelevant, but a great turnover will be generated. So, when there is a high value density, you can
use a centralized system and not consider efficiency in transportation costs. On the other hand,
Coca-Cola has a low value density, since unit value is low, but weight is quite relevant, so
transportation costs are very high. In this case, an efficient transportation system should be
implemented, and therefore, a decentralized system as well. To sum up: In centralized systems, also
if you spend a lot on transportation, you can afford them if you manage high value products,
otherwise choose decentralization system.
à Why do many companies prefer to use a centralized system, even though transportation may be
high? In order to understand this reason, we consider the following example: here you see the map

74
 of Europe, and we consider a company in EU that
serves UK, Italy and France, which has a centralized
system. It has a unique warehouse that offers
products to all the three countries. The expected sales
for the future are 100 in each country, for a total of
300; therefore, 300 pieces in the warehouse are
stocked. However, once the WH receives the actual
orders from clients, the company realizes that Italy
has an actual demand of 50, France 50 and UK 200. In
this specific case, this is not a problem, because the
stocks are all in just in one warehouse, not in several
distribution centers around the world; the total
amount of demand is still 300, so the company can still satisfy it, it just has to rearrange the
distribution of stocks of the three countries. This would be completely different if the company had
a decentralized system, with several distribution centers, one in each of the States of interest. In
that case, firstly products are moved from the WH to DC, so some transportation costs are incurred
and then, due to the actual demand, if all of them have 100 initially, Italy and France have to give 50
of their units to UK in order that the latter can reach 200. Therefore, new deliveries have to be
planned and paid; other transportation costs are incurred again.
Companies prefer centralized system to cope with variability of the demand, since the latter is more
and more volatile nowadays, in almost all industries. Centralized system is therefore, more suitable
when there is high volatile and high density products.
o Postponement/speculation: this decision refers to 1) what we have stock (what should be placed in
the warehouse: a finished product or a semifinished product?) and 2) where to stock (either in the
warehouse close to the factory or in the local DCs). To understand this decision, two concepts have
to be introduced: Speculation means to carry out an activity on the basis of a forecast concerning
customer’s future orders and Postponement means to carry out an activitiy later in time, i.e. once
the order from the client has been received.
-Postponement can be applied in manufacturing and logistics.
Manufacturing postponement: the product is manufactured or completed on the basis of an actual
order from the client. Semi-finished products are stocked.
Logistics postponement: the product is moved downstream in the logistic network when the
company receives an actual order from the client. Stock of products is held in central WH.
-Speculation can also be applied in manufacturing and logistics.
Manufacturing Speculation: the product is manufactured and completed on the basis of a sales’
forecast. Finished Products are stocked.
Logistic Speculation: the product is moved downstream in the logistic network on the basis of a
sales’ forecast. Stock of products is held in DC.
Postponement/speculation matrix: logistics can be
applied either through stock in DCs or in WHs (far
away from customers), while speculation can be
carried out either through the make to stock or make
to order approaches. By combining these situations, I
can get four different combinations:
1) Speculation in logistics, so stock in DCs and
speculation in manufacturing, so make to stock:
in this case the warehouse is close to the factory,
and there are the distribution centers. The stock is
represented by the triangle, and therefore it’s
placed in the DCs. Because I’m using make to stock, I’m stocking the FP, so finished product.
Characteristics of this system: high speed, high holding costs (we’re stocking finished products

75
 that have, compared to semi-finished products, high unit value and moreover, a lot of pieces are
stocked to cope with volatility of demand), low transportation costs (movement of finished
products from the WH to DCs based on a forecast, through an economy of scale, so by moving
high quantity to reduce costs) and low production costs, since we use make to stock approach.
Example: PPG is a company that provides paint for, cars, which produces in Milan and provides
FP to all warehouses around Europe. Since paint it’s a product of low density, they use a
distribution system which is organized with a number of national DCs and they work with a
make stock approach, so the FP is stocked in the national DCs, in order to guarantee fast
delivery. If instead they would have held the FP in the Milan WH, of course the delivery would
have been slower.
2) Postponement in logistics, so stock in WHs and speculation in manufacturing, so make to stock:
we’re stocking the finished products, since we’re using make to stock, in the WHs.
Characteristics of this system: medium speed (now the product is not stocked next to the final
customer, but at least the product is a finished product), medium holding costs (stocking the
finished product is expensive, but the amount of stock hold is lower), high transportation costs
(products to be delivered to a regional DC are put together, but this procedure is conducted
frequently) and low production costs, since we’re managing production with a make to stock
approach.
Example: Coca-Cola has a volatile demand, especially in the short-term, with low density.
Therefore, it’s quite risky to have products in the distribution centers. You cope with a situation
like this by holding all stocks in the WHs and once the orders are received, from a specific
region, the company will organize the order for that region and deliver to the regional DCs, that
consequently will organize the shipment to final customers.
3) Speculation in logistics, so stock in DCs and postponement in manufacturing, so
make/assemble to order. In this case, semi-finished products are held in the DCs.
Characteristics of the system: medium/high speed (the product is stocked next to the client, but
it’s a semi-finished product), medium holding costs (a lot of pieces are stocked, by since they are
semi-finished products, they have a lower unit value), low transportation costs (deliveries of
large quantities from the warehouse to DCs are made on the basis of a forecast) and high
production cost (make/assemble to order is made to offer customization, which require higher
production costs).
Example: IKEA, since the stock is held in DCs, which in this case are the stores. If you buy
furniture, you find the pieces to assemble, so a semi-finished product.
4) Postponement in logistics, so stock in WHs and postponement in manufacturing, so
make/assemble to order. In this case, semifinished products are held in the warehouse, so
products have to be customized, but this procedure has to be conducted in the warehouse, not
in DCs.
Characteristics: low speed (product is far away from the customer and it has to be completed),
low holding costs (semi-finished products have a low unit value and just a few pieces are held,
since you’re centralizing stock in one place), high transportation costs (items are delivered one
by one) and high production costs (customization is costly).
o Outsourcing: who should manage warehouses and distribution centers? Outsourcing it is becoming
more frequent, especially due to the cost and the complexity of the activities that warehouse, and
transportation management, involve. Moreover, outsourcing is common due to technicalities, due
to legal constraints. Nowadays, more and more frequently value-added activities are carried out by
3 parties logistics providers. They can concern:
§ The physical flow of products and materials
§ The information flow.

76
 Therefore, when you decide to outsource these activities, you
have to be sure that the logistics providers chosen is the best
one.
Value added activities on physical flows: an example of this may
be a factory in Lombardia that produces socks, which of course
can easily forecast demand in long-term, but in the short-term
it’s not easy to be deducted, especially it’s not easy to
understand how many products will go to the supermarket and how many to the retail store. In fact,
this company provides products to retailers and supermarket chain: the product is the same, but at
the same type packaging is different (in the store, there is no packaging, in the supermarket, there
is). Volatility of demand can be solved through the postponement of packaging, so the company
stocks in the warehouse semi-finished products, so without packaging, and just once the order is
received the packaging is provided. In the recent years, however, the company decided to outsource
the management of the warehouse to a third party, so also the packaging procedure is relied to the
third party. Therefore, the logistic providers is in charge of the value added activity carried out on
the product – consequently, it’s fundamental to carefully select the best logistic provider.
Value added activities on information flow: information usually
refers to orders coming from the clients. In this case, let’s consider
Nestlè, which selected a logistic provider to manage delivery in Italy,
especially of Nespresso items (pods – cialde). Pods can be
purchased online or in a store, but due to the complexity of the
home delivery, Nestlè identified in Milan a logistic provider
managing the warehouse activities, the transportation and the
orders accomplishment. Outsourcing an activity like this, it’s important to select that the logistic
providers treat data in a private way and that is able to manage these activities.
3) In the pyramid functional and implementation and operation choices are very technical, so we don’t really
need to talk about them.

lesson 15
12.03.2021

COLLABORATIVE PRACTICES:
We now analyze how companies can cooperate with each other to improve their supply chain performances. Vendor
Managed Inventory, Consignment Stock, Continuous Replenishment Programs, Collaborative Planning, Collaborative
Forecasting and Collaborative Planning, Forecasting and Replenishment.

Vendor Managed Inventory

How is inventory management fulfilled in a normal company that does not collaborate?
Let’s say we manage stock, and we can use several techniques in order to do it. For example, we can use a technique
that is based on the ROP.
Let’s say that horizontally we have time and vertically we have pieces, we place an order to the supplier at the
beginning of the year and we receive the delivery exactly here (--) and day by day we’ll keep picking items from the
stock and this means that the stock level will decrease over time. Sooner or later, we’ll reach a moment when we
realize we need to make an order to the supplier and when do we reach this moment? There are different techniques
in inventory management. One of the most common is a technique in which we decide when to place an order on the
basis of 2 dates which are:
• the delivery time of the supplier (let’s say 5 days)
• demand rate à number of pieces on average we will need on a daily basis (let’s say 2 pieces every day).

77
Now the question is: when should we place any order to the supplier? During the delivery time we’ll need two pieces
a day so if we reach a total amount of stock that is equal to 10
pieces, for sure this is the moment we’ll need to place the
order. Why? Because 10 is the number of pieces which is
large enough to satisfy the demand during the delivery time
pf the supplier. If you place the order later for sure, we’ll go
out of stock. When we place any order, we do so when we
reach the so-called reorder point which in this case is equal to 10 pieces. We know that when we reach this specific
level of stock, we’ll need to place an order to the supplier.
Now let’s say we place an order to the supplier in a time in which he has lots of orders to fulfill so in this situation
there’s a high probability that we will not receive the order in 5 days because he’ll be late, and we’ll go out of stock.
We also have to consider that the supplier is exposed to some risks: for example, we are a retailer (supermarket) and
we know that in the next week we’ll have a promotion for a product so here we are dealing with the stock of
spaghetti, we need to place an order and we know that during this specific week we’ll need a higher number of stock.
We place an order higher than the normal one. The supplier may not be ready for that amount. Why? Because the
supplier is organized on the bases of the standard level of consumption of the costumer. So, we are exposed to this
kind of risk, since the supplier may not be able to meet our need.

Inventory management is not perfect in companies, because it is managed without a real collaboration with the
supplier. When you place an order to the supplier, he receives the demand and only in that specific moment he will
acknowledge the quantity you request. This exposes the company and the supplier to many types of risks. There is an
alternative: Vendor managed inventory which is a practice according to which the vendor (supplier) will manage the
stock on behalf of the costumer. In order to do so, the costumer must provide some information to the supplier, like
current inventory level and expected consumption of the product in the near future. They also have to agree some
things on a contractual basis, like:
• Maximum & Minimum amount of stock at the customer’s site
• Service level to the customer: percentage of time in which the client doesn’t go out of stock.
• Frequency of update of the information

Example: inventory management system in which:

à horizontally: time
àVertically: pieces
According to inventory management the client will inform the supplier about its current inventory level. The client will
inform the supplier about the evolution of his inventory and the supplier is able to observe the evolution of the stock
level of the client with given frequency. Frequency has to be agreed upon in order to implement this system.
Generally, the frequency usually is a real time information and thanks to this, the vendor is able to see moment by
moment what the stock level of the client is.
The client and the vendor had agreed on a minimum and maximum level: the
client wants a minimum level since he doesn’t want to go out of stock so he
wants to make sure that in any moment I will have a minimum amount of
stock. Also, a maximum amount will be agreed because the client doesn’t want
to receive too many pieces from the supplier.
It will happen that the client will consume day by day some pieces and again it will be that in a day the stock level will
reach an amount quite close to the minimum level – and it can be, that in this moment, the vendor knows that the
stock level is almost equal to the minimum level and that in the next three weeks the customers won’t use any pieces
of this item. Therefore, for three weeks the stock level will be held constant, and at the end of the three weeks, when
the client informs that the vendor that the demand will increase, so the supplier makes a delivery. This process is
repeated again and again.

78
What are the reasons why the supplier and the client should use the vendor managed inventory?
The benefits for the customer are:
• Outsourcing of an activity that requires resources: inventory management is outsourced.
• Higher service level provided by the supplier
• Lower holding costs, due to the fact that the average inventory level is lower with vendor management
inventory
The benefits for the supplier are:
• Easier and more effective production planning process
• Lower stock-out costs
• Lower holding costs, due to the fact safety stock to cope with demand uncertainty is not required anymore
(or is much lower)

Why do you think within the vendor management inventory the service level of the supplier is going to be higher?
What is the reason? Because the supplier has more information to better organize deliveris. What are the information
that let the supplier organize in a better way the deliveries? à it concerns the expected demand from the client. In a
normal system if there is no collaboration, the supplier will wait for the order from the client. In roder to be ssure that
delivery can be done, the supplier carries out demand forecatsing processes and on the basis of the forecatsed
demand the supplier will organize its own activities so when the order will arrive he will be ready to make the
delivvery. This can be risky because demand forecast are always wrong.
What happens with vendor managed inventory? If you are the supplier you will recevie the expected demand direclty
frm the client and this information will be usefull. In this case we expect a high service level.

Why do we have a lower holding cost? the holding cost is the financial cost we incurr because we have stock in the
warehouse. We have a lower stock because we can decreae the amount of the safety stock which is amount of pieces
we put into the warehouse and it will be use in case of an urgency or unexepcted event. We use safety stock to cipe
with risks like delays in the delivery time, risk of an unexpected increase of the demand of the client. If you are in
situations like this you will have a higher holding cost.
With vendor managed inventory you can decrease safety stock. Why? Because that the risk of stock out of the
supplier is going to be lower.

Why should the supplier be happy to work according to this practice with the client? à he gets many benefits (see
picture up).

Consignment Stock
This practice is the same as VMI. The only difference refers to the fact that the ownership of the products goes from
the supplier to the customer only when the customer picks the product from the warehouse. In this case, the product
is owned by the supplier even if it is in the customer’s warehouse, until the customer needs to use it – the client
doesn’t have holding cots. The customer can postpone to the latest possible time the payment due to the supplier.
The benefit for the customer is generally counterbalanced with better financial conditions for the supplier (earlier
payment conditions etc.). Furthermore, the risk due to any accident in the warehouse is taken by the customer and
not by the supplier.
For example, instead of paying within 60 days the supplier, the payment can be done within 30 days – in this way, the
negative effect of consignment stock, with a good payment agreement.

Continuous Replenishment Program

It is similar to VMI, but in this case the supplier proposes a replenishment to the
client, which can be done only after it has been authorized. When suppliers
understand that is time to make a replenishment order to the client, he makes a
proposal: can I deliver tomorrow? The client can either refuses it or accepts it. This practice is very common between
fast moving consuming goods (Barilla, Ferrero, Nestlè) manufacturers and big supermarket chains or department

79
stores. If you’re Esselunga you may have multiple suppliers, by chance in the same day all suppliers rearrange the
delivery and this may be a problem, so you organize by postponing and rearranging all the replenishment orders from
the different suppliers.

Collaborative Planning
According to this practice, production plans are shared among the main players of the pipeline (manufacturer,
suppliers, subcontractors). In case one of them has to change some decisions concerning future requirements or
production quantities, this variation is communicated to the partners so as to check whether they can support it.

Collaborative Forecasting

It aims at leveraging the knowledge about the market demand present in different layers of the supply chain in order
to produce the best possible demand forecast. It aims at producing one single demand forecast (“single number
forecast”) for the entire supply chain.
Example: there are three players; Barilla is the manufacturer, Esselunga is the retailer and then we have the customer.
Barilla and Esselunga make their own forecasts, which may not be equal to each other. In order to overcome these
differences, players cooperate with each other to develop jointly the demand forecast, to obtain a single number
forecast. With collaborative forecasting, a team of forecasters have to be composed, with a leader: in this setting, the
leader should be Barilla, to have a view of the bigger picture, since the bigger market it serves.

Collaborative Planning, Forecasting and Replenishment (CPFR)

It is a form of collaboration among two or more operators of the supply chain, which jointly carry out and plan the
following activities:
- Promotions
- Forecasting
- Production/Procurement Planning
It’s a very complex situation, therefore it’s not widely spread.

80