FMC Biscuits Plans Production and Logistics Allocations for its Supply Chain1

A large FMCG company, FMC Biscuits (FMCB), makes and sells a wide range of biscuits,
cakes, certain dairy products, breads and other baked food products. It is a leading national level
company and its business is spread all across India.

Supply Chain

FMCB produces all its products in 45 plants. 260 SKUs of biscuits, cakes and other food items
are produced from these plants and transported in trucks to depots from where they are supplied
daily to wholesale or distributor/large retail points. Out of a total 66 depots in India, spread
across all states, 8 are larger depots (called ‘mother depots’). The mother depots supply SKUs to
both wholesale/distributor/retail facilities and other depots.

Markets and SKUs

While some brands of FMCB are strong, there is stiff competition from both large national
players (like FMCB) and local brands. The SKUs are sold from all types of retail outlets – big
and small. Various retail outlets are classified as ‘modern’ or ‘traditional’ trade. Modern retail
comprises organized and relatively large retail stores and chains like Big Bazaar, Spencers,
Reliance Fresh, More, etc. Traditional retail consists of small Kirana shops which dot the roads,
streets and corners of urban and rural India. While distribution and sales to Modern retail stores
is better organized, distribution to Traditional stores is carried out by local distributors. FMCB’s
salespersons visit several Traditional stores every day and collect SKU-wise orders for the next
day and intimate the local distributor to deliver them. The distributors in turn place orders on to
nearest depots.

The 260 SKUs are grouped into 31 Pack Types and 36 Brands. The biscuit SKUs are also
categorized as ‘cream’ and ‘non-cream’. While the SKU level daily sales for a distributor shows
high fluctuations, Brand level sales figures are relatively less fluctuating. The practice is to
forecast Brand level sales at a depot, which is easier to do. These forecasts are a key input to the
process of planning of Plant to Depot dispatches and Plant level production.

Production

Out of 45 plants dedicated to FMCB, about 25 are sub-contracted and the rest are self-owned.
The investment capacity and quality of process equipments (production lines) in self-owned
plants are at a distinctly higher level than in the sub-contracted plants. The premium SKUs and
the ones which require special quality control effort in production are made in self-owned plants.
All 45 plants make biscuit SKUs, whereas the other categories of products (like cakes) are made
in fewer plants (typically about 10-25).

A plant location has 2-6 production lines, some for cream biscuits, some for non-cream biscuits,
and others for cakes and other baked items. A production line, depending on its technology, can
1
Prepared by Rahul Pandey, Sridhar Vallala and Amiya Kumar Nanda in 2016, [email protected]
produce from a given set of SKUs at a certain output rate. Therefore every production line is
shared among several SKUs. Capacity utilization of a line depends on the number of SKUs it
produces (as more SKUs imply more changeovers, especially the change from one category of
SKUs to another category involves longer time to changeover).

The Head of Production prefers that a line, while being utilized well, must produce SKUs from at
least three Brands over the course of a month.

Outbound Logistics

Large trucks – of two sizes – transport SKUs from Plants to depots or to mother depots. In order
to keep the freshness of stocks the Head of Supply Chain prefers to transport majority of the
SKUs from Plants to depots within the same region. This is especially so for the SKUs which
have high volume sales within a region. For this purpose India is divided into four regions –
North, South, East, and West. Each region has several Plants and depots. Each region has at
least one mother depot too.

However, certain Brands which are high value (premium) brands with a good presence in all
regions, and with high margin, are designated as ‘National Brands’. A lot of such ‘national
brands’ have lower volumes of total sales as compared with the ‘mass’ brands. The National
Brands are produced centrally in one or a few plants in the country as they require certain special
operations in production and have lower total volumes. Hence these Brands need to be
transported from one region to another.

In addition, if demand for some SKUs in a month in all depots in a region surpasses that region’s
production capacity, the gap is often filled in by transporting from another region if the latter has
surplus production capacity at Plants or from excess inventory stocks at some of the mother
depots. However, in order to reduce logistical inefficiency the Supply Chain Head has mandated
that certain Brands in a depot should not be sourced from more than three Plants. In addition, the
logistics teams stationed at Plants make sure that outbound trucks are loaded to a good level of
utilization.

Network Operations-Logistics Allocation Planning Issues

FMCB carries out a supply chain-wide network operational allocation planning every month for
a horizon of six months. The plan made includes month-wise and SKU-wise plans of production
at various plants (factories) and of logistical allocations from plants to depots. These plans are
made for six months in order to consider inter-month seasonality of demand and accordingly also
prepare end-of-month carryovers of stocks. First month’s plan is frozen for operations (i.e.
production and outbound transportation plans) whereas subsequent five months’ plans are
conveyed to the procurement department and production department to plan for longer-term
procurements and resource deployments.

The objective of this monthly exercise is to meet forecasted demands in all the markets served by
the depots from the existing production line capacities at the plants so that total variable cost of
production and outbound logistics (including transportation and taxes) is minimum at the all
India level. At the same time the management wishes to guarantee a minimum service level to
the depots in terms of the lead-time responsiveness of distribution from plants (or mother depots)
to final depots.

There are several variable costs that are proportional to the quantity of production or
transportation, like variable cost of production (conversion, fuel, labour, etc.), and cost of
transport and handling. Transport and handling rates are region (or route) specific. Many of the
taxes and duties like VAT, CST, Octroi and Excise are also location specific. Some states offer
VAT exemption if Plants of those state supply to within-state markets.

On the outbound logistics front, the mother depots have larger storage space than the depots.
The mother depots supply to both their neighbouring distributors and to other (smaller) depots.
The former can therefore bring in economy by taking in larger lot supplies from plants (by bigger
trucks), by holding larger stocks, and by allowing the possibility of aggregating safety stocks of
multiple smaller depots. But the mother depots are less responsive than the smaller depots as the
former are located far away from most markets. In order for a mother depot to be economical, it
must get supplies at least once a week by a truck of not less than 16 tonne capacity.

A senior supply chain planning executive had suggested the following approach: Calculate the
least total variable cost (or cost-to-serve) Plant-Depot route (labelled L1 route) for every depot;
Start allocating shipments from L1 routes for the depots, starting from the first depot in the list;
Whenever there is inadequate capacity left at a Plant for supply to a Depot by L1 route, switch to
the next least cost route (L2) for that Depot for the remaining quantity. After all allocations have
been made, incorporate the remaining constraints of logistics, sales, production, etc. by make
appropriate adjustments to the allocation numbers.

This logic appeared simple and was easy to execute on a spreadsheet. However, it involved
human judgment at several steps. One planning executive differed from the other in the priority
of reallocating quantities to L2 routes and making adjustments while incorporating various
constraints and preferences of logistics, sales and production teams. In addition, the supply
chain planning team was not sure if the calculated plan was really the least cost one. A few
executives also often raised a concern that some of the constraints and preferences of sales,
production and logistics teams were not fully addressed in the plans, which resulted in the
planning team having to make many changes on a day-to-day basis as the production and
dispatch quantities executed by operations executives were often different from what was
suggested at the start of a month. Some even claimed that there was too much uncertainty,
especially in the markets, and so it was simply not possible to freeze a plan in advance.

Presentation Questions
What are the challenges faced by FMC Biscuits while planning its network allocations
(from plants to depots)?
How would you scope and structure the problems of supply chain network
allocation? (try to answer the questions below)

Discussion

1. What trade-offs are involved in the monthly network operations-logistics allocation

2. An initial data collected by a consultant is provided in the accompanying Excel file

3. Can you build a model to help evaluate alternate plans and suggest an optimal plan?

a. What are the key decision variables?

b. What is the objective?

c. What are the main constraints?

(Without using full data of the Excel file can you illustrate you approach by assuming a
small dataset? If you do not have time to code the model, then try to explain its
formulation and logic.)

4. What kinds of outputs can you provide to different executives (production/factory heads,
logistics heads, sales heads, top management) that’ll be useful for their decision-making?
What kind of operational and tactical/strategic analysis is possible with the above model?