openSAP

Planning with SAP Analytics Cloud

Week 1 Unit 1

00:00:05 Hello, and welcome to the course. My name is Holger Handel, and I'm a product manager for
SAP Analytics Cloud.
00:00:13 I'm focusing on the planning capabilities of the tool. Many of you might know Analytics Cloud
00:00:18 already from BI or analytics projects, but few people actually know that Analytics Cloud offers
also very powerful planning capabilities.
00:00:26 This course focuses exclusively on planning with SAP Analytics Cloud. So let's have a look at
the detailed agenda.
00:00:33 We will provide five weeks of intensive training, covering all the relevant aspects of planning with
SAP Analytics Cloud.
00:00:41 We will start in week one with the core planning capabilities. So you will get to know the concept
of collaborative enterprise planning
00:00:49 and then you will see all the relevant planning features in an extensive demo. After this, we will
turn to the core planning concepts so you will see how you
00:00:59 can use the modeling environment to create data models, and you will also see then the
standard planning features in the story,
00:01:05 like, for example, the version management and the planning panel. In week two, we turn then to
the advanced planning features.
00:01:13 so planning is typically more than just manual data entry, normally, you also want to implement
complex calculations.
00:01:21 SAP Analytics Cloud offers different concepts for this. So we will see the value driver tree
component, but also the data actions,
00:01:28 the advanced formulas, and the allocation engine. In week three, we cover then more the
process- related topics.
00:01:36 So typically planning is a process that involves several people of the organization.
00:01:42 So with a planning calendar, SAP Analytics Cloud offers a powerful component to orchestrate
and manage the process.
00:01:49 Week four is dedicated to the topic of predictive planning. With smart predict, Analytics Cloud
offers a very powerful environment
00:01:57 to create and train your predictive models. You can now also seamlessly integrate
00:02:02 these predictive models in your planning workflows. In week five, we cover the topic of custom
planning applications.
00:02:10 With the Analytics designer, you can create your own planning applications and we will also
introduce the SAP Analytics Cloud add-in
00:02:19 for Microsoft Office, which is basically a new Excel add-in. Finally, in week five,
00:02:25 we will also see best practices for user experience in planning applications. As with the usual
openSAP courses, we will have a self-test after each unit so
00:02:37 you can basically check your understanding of the covered materials. After each week, there will
be a weekly assignment,
00:02:44 and then finally in week six, there will be the final exam, which is a prerequisite then for the
certificate of achievement.
00:02:52 In addition, we also provide here a community where you can exchange with the participants,
but also with the instructors of the course.
00:03:02 With this course, we want to basically target the expert users for business planning in an
organization, but also, of course,
00:03:09 consultants who need to implement SAP Analytics Cloud for their clients. Beyond that, of course,

00:03:15 we are open for everyone who's interested in planning with SAP Analytics Cloud. The course
itself has no hard
00:03:23 prerequisites, but of course, it might be beneficial if you have already had some exposure to
topics in BI financial planning,
00:03:30 or also in the predictive area. Some background in existing planning tools like, for example,
00:03:35 BPC might also be beneficial, but are also not a prerequisite for this course. For this course, we
will also provide additional hands-on exercises,
00:03:45 which will intensify your learning experience. The guides and additional content will be
00:03:51 provided as additional downloads on the course Web site. The hands-on exercises are not
mandatory for obtaining the certificate of achievement,
00:04:00 but we will highly recommend doing these exercises because we will cover then in some of the
aspects in more detail.
00:04:07 With this, I want to finish today's unit and I hope to see you next time again. So, thank you and
goodbye.

2
Week 1 Unit 2

00:00:05 Hi, thank you for joining today. In this session, we will cover the main concepts
00:00:10 of collaborative enterprise planning with SAP Analytics Cloud. Before we do that, let's have a
look at the overall architecture.
00:00:17 As you can see, SAP Analytics Cloud provides business intelligence, planning and predictive
capabilities all in one place.
00:00:25 With the Analytics designer, users can create additional custom applications by using APIs and
SDK extensions.
00:00:32 SAP Analytics Cloud can connect to a various number of enterprise systems, on-premise data
sources and cloud data sources.
00:00:40 Let us now focus more on collaborative enterprise planning from a business perspective.
00:00:46 For many companies, the strategic and financial plans are a central piece within their planning
solutions.
00:00:53 The strategic plan sets the company's overall direction and targets on an aggregated level by
using
00:01:00 planning assumptions, growth drivers, and KPIs for the next years. On the other side,
00:01:07 the financial plan with the profit and loss statement, balance sheet, and cash flow is one of the
main results of companies' planning solutions.
00:01:15 These plans are surrounded by the operational plans inside the organization, like sales,
workforce, supply chain, IT, marketing, and more.
00:01:25 All those plans need to be connected as they are affecting each other. As an example, let us
focus on sales planning.
00:01:32 While there might be a sales target that is coming from the strategic plan, we will probably plan
some quantities and prices.
00:01:39 As a result, we will get a revenue number. The planned revenue then needs to be part
00:01:44 of the profit and loss statement in the financial plan. So as you can see,
00:01:48 collaborative enterprise planning is about a various number of operational and financial plans
that are integrated with each other.
00:01:56 How can SAP Analytics Cloud support companies in addressing all those different business
requirements?
00:02:01 The good news is that SAP Analytics Cloud provides a powerful toolset that companies can use
to create and connect different plans.
00:02:10 At the beginning, we need to create a planning model. The model consists of dimensions
00:02:14 and hierarchies that we need as a structure for our data. So now, as we have this empty box,
00:02:20 we can load master and actual data from the underlying source systems as a starting point for
our next planning cycle.
00:02:29 To set up a planning process, we need a workflow support component. In SAP Analytics Cloud,
we therefore can use the calendar.
00:02:36 Here, we can create a process that contains tasks. Every task can have an assigned planner
that is responsible for this
00:02:43 particular planning step, and a reviewer that will control and accept the planned values. When
the process has been started, all planners will be automatically
00:02:51 informed by the system with their needed actions. In most cases, we do not want to start with an
empty plan.
00:02:58 To prefill this plan, the easiest way is to copy our latest actions into the plan, maybe using some
growth drivers.
00:03:05 As many companies tend to have more sophisticated requirements these days, we can also use
predictive capabilities to create an automatic planning proposal
00:03:13 that then can be adjusted by the planners. During the planning process,
00:03:18 users can then input and adjust data at any level. The system will automatically disaggregate
plan data to the lowest level.

3
00:03:25 To support the users within the planning process, SAP Analytics Cloud offers out-of-the box
distribution and spreading functions.
00:03:33 And for more complex business requirements, users can create allocations and advanced
formulas using script logics.
00:03:40 When the planning process has been done, users can directly review their plan data by
comparing the plan with the latest
00:03:46 actuals or the latest forecast with the initial plan. As you can see, the combination of BI,
00:03:52 planning, and predictive are coming all together in SAP Analytics Cloud. At the end of the
planning process,
00:04:00 we now can finally submit plan data, set data locks so that no one can change the final version
anymore, and forward the data wherever we need it next.
00:04:10 During the whole process, users can collaborate with each other. They can share stories, enter
comments,
00:04:16 and run discussions along the overall planning cycle. With this, we have to scrap the first
00:04:22 pillar of collaborative enterprise planning to crowdsource plans and budget. With this toolbox, we
can create various
00:04:28 plans with different business requirements like finance, sales, HR, and more. As we know from
the beginning, a key aspect of collaborative enterprise planning
00:04:38 is the integration between the single plans. Therefore, we can align and link
00:04:42 the operational and financial plans across the organization. One more example here is related to
investment planning.
00:04:48 Think about investing in a new building. The new factory will increase your assets inside the
balance sheet,
00:04:54 while the calculated depreciation will be part of the profit and loss statement. Last but not least,
the planning solution needs to be connected to the surrounding
00:05:04 enterprise systems within the organization. As an example, we can load master data and
transactional data from S/4HANA
00:05:11 into SAP Analytics Cloud, and we also can write back data from SAP Analytics Cloud directly
into the ACDOCP table inside S/4HANA.
00:05:20 Or we can load data from SuccessFactors into SAP Analytics Cloud as the basis for HR
planning.
00:05:27 To summarize, SAP Analytics Cloud supports collaborative enterprise planning with a powerful
toolset to create a number of various plans.
00:05:36 The single plans can be integrated with each other to improve the overall alignment, and SAP
Analytics Cloud can connect
00:05:45 to many SAP and non-SAP data sources across the organization. Thanks for watching, and see
you in the next session.
00:05:52 Goodbye.

4
Week 1 Unit 3

00:00:05 Hello, everyone, and welcome to the openSAP course on SAP Analytics Cloud Planning. In this
session, we will show you a demo
00:00:13 on collaborative enterprise planning, which will cover the process of monthly forecasting, where
we also include live data from S/4,
00:00:23 as well as a multi-year strategic plan, which is then derived over into an annual budget plan
where we also leverage the functionalities and the integration
00:00:34 of smart predict in SAP Analytics Cloud. Enjoy watching.
00:00:39 SAP Analytics Cloud planning is the platform that unifies planning with augmented BI and
predictive all in one solution.
00:00:47 In the demo, we will demonstrate our collaborative enterprise planning capabilities for a bike
manufacturing company at the end of Q1.
00:00:56 We will start with the review of our overall company results. From there, we will do a forecast
until the end of the year,
00:01:03 then go into strategic planning and create a three-year plan, which will then cascade down into
the organization to create the annual budget.
00:01:12 In addition to global controlling being involved, we will incorporate two additional LoBs into the
annual budget process
00:01:19 for the marketing campaign planning and the HR workforce planning. Let's get started.
00:01:28 As a company that produces and sells bikes, we will start with a dashboard giving us an
overview of the main company KPIs, based on Q1, covering sales, finance, workforce.
00:01:39 Our overall revenue seems to be good. The geomap is visualizing our revenue situation
00:01:45 across the globe and is showing in red which regions are below budget. Additionally, we see the
workforce KPIs
00:01:53 predicting a high attrition rate for the current year, and the last employee survey score is
concerning, which is being explained
00:02:00 by the comments left via the commenting widget. Let us ask the system for the most relevant,
hidden insight into the numbers
00:02:09 by using the SAC augmented BI feature called Smart Insights, which will apply machine learning
capabilities on our data.
00:02:17 We can right click on the variant between actuals and budget to access the Smart Insights
feature,
00:02:23 which works out of the box on any SAC model to get the system to provide us with the top
contributors that make up this variant.
00:02:31 We now easily see that the e-bikes are our top revenue growth driver and that the traditional
bikes are seeing a decline.
00:02:40 The market seems to be shifting to the e-bikes faster than we expected. The revenue breakdown
page.
00:02:46 We see our revenue numbers in more detail along our regions and products. Looking at the
table in-cell charts,
00:02:53 we quickly find that the mountain product line is showing the largest decline across all product
groups.
00:02:59 And we also see that the mountain group does not have any electric bikes in its offering.
00:03:08 Starting with the revenue forecast, I will, first of all, create a brand-new version as a copy of the
actuals.
00:03:16 By simply right clicking on the actual version directly in the table and selecting Copy Version,
00:03:23 I will create a new forecast version and call it Smart Prediction. The system automatically adds a
new version to the table.
00:03:33 I see that the version is empty from Q2 until the end of the year. Same as my actual version.
00:03:41 Next, we go into a smart predict functionality, which will utilize machine learning capabilities to
automatically create

5
00:03:47 a time series forecast of our revenue until the end of the year. We start off by seeing the
overview page of a previously created predictive model,
00:03:55 the relatively low MAPE value, which is the key performance indicator and stands for mean
absolute percentage error, basically tells the difference
00:04:03 in percentage between the actuals and the model predictions. The lower the MAPE, the better.
00:04:09 Here we see several models we already created in the past. This features a self-serving
business
00:04:15 user capability and does not require any programming or deep data scientist expertise. I can
quickly create a prediction on my own, so let's do it.
00:04:25 As this feature works on any planning model out of the box, I'm able to simply select my demo
model called Living Company.
00:04:37 You can choose any version storing our historic data as well as our most recent actuals.
00:04:45 Our signal variable is the account or the measure that we want to forecast. In this case, it is our
sales revenue.
00:04:54 For the time series forecast, we need to also select our date dimension. Finally, we define the
segmentation and choose the region and product dimension.
00:05:04 What this means is that the system will create one time series forecasting model per region and
product combination in order to achieve the most accurate
00:05:12 result by working on the detailed data level. As we want to create our forecast until the end of
the year,
00:05:21 I need to choose nine periods to be created. Now that everything is configured,
00:05:26 we can click on Train, and the system will start running its predictive algorithms. Once the
training is complete,
00:05:41 we will get a detailed summary presented of the results. In the Signal Analysis,
00:05:48 we can differentiate an overall trend, which in our case is positive, decoupled from the cycles
which might reflect seasonality or outliers.
00:05:57 Now that we gain the necessary confidence in the predicted result, we will write it into our newly
created version we called Smart Prediction.
00:06:12 Let's now go back into our Revenue Forecast page and look at the predicted forecast in our
table.
00:06:19 With a detailed result created across all region and product combinations until the end of the
year.
00:06:25 Comparison to the previous forecast shows a decline, and we see the new forecasts predicting
us ending the year below budget.
00:06:37 With the revenue forecasts concluded, we can now move on to the expense forecast. The
calendar component in SAC is the inbuilt workflow component that helps
00:06:48 us create workflow for various planners and reviewers. This case I'm going to assign
00:06:52 to the region Germany, and a story has been attached to the task. The task also carries the filter
context for story tool.
00:07:00 As we enter the story, we see that the filter has been applied to Germany. As we are processing
our forecast for April, the cutover date is set
00:07:09 accordingly, and we see the remaining months as input ready in the table. Due to the current
situation,
00:07:17 we will forecast 40 percent reduction of our building expenses for April. Next, I want to focus on
the travel expenses,
00:07:27 which are also heavily impacted by the current situation and the travel freeze, which has been
implemented company wide.
00:07:35 If we scroll down, we see two additional tables, which are showing live data from our S/4HANA
system.
00:07:43 This table is giving me a live comparison of my SAC forecast plan, with the very detailed travel
expense line items from the S/4HANA systems.
00:07:54 This is achieved using our live planning feature, which does not require any data replication

6
00:08:00 and is enabling me a view into of my transactional system, including much more details behind
the travel expense account than I have
00:08:08 in my planning model, the live blending of planned data with other data sources works with all
live connectivity supported systems.
00:08:16 A few examples are S/4HANA, BW, SuccessFactors. Now, I easily see that my forecast in April
is 115,000 above my current actuals.
00:08:28 With the travel freeze in place, I decide to simply copy the latest actual value into my April
forecast.
00:08:35 We see the blender table is also updated and is now showing that the actuals and the plan in
April are identical.
00:08:42 Once finished with the forecast numbers, we will submit our task. Using the value driver three,
00:08:57 we will be able to create several versions, using different business drivers and evaluating their
impact on the topline as well as the bottom line.
00:09:06 Let's do some additional adjustments on version one. I will increase the investment into our
mountain bike product group
00:09:20 by raising the revenue targets for the year 21 and 22. As I change the growth rate,
00:09:31 I immediately see what this change means for the overall revenue and the delta to the current
baseline, via SAC on-the-fly calculation capabilities.
00:09:48 This, in turn, will also add that my expense budget, based on the calculations I have modeled
00:09:53 individually, I see the sales and marketing production and the G and A expenses now adjusted
as well.
00:10:02 This should give the company the necessary resources to expand the mountain bike portfolio.
00:10:07 We also see a slight increase in the overall operating income. Additionally, to tackle our high
attrition problem we saw on the KPI page,
00:10:17 I will increase the employee bonus and incentives to 20 million for the next two years.
00:10:31 This will slightly lower my operating income. Using the waterfall chart on the right, I can compare
the two versions and will
00:10:41 make the decision to go forward with version one as our future strategic plan. The operating
income is copied
00:10:50 into the budget by simply using the copy and paste functionality. The system automatically
performs a deep copy and assigns the values
00:10:59 to the P and L accounts and the quarters or months with the same breakdown as available in the
strategic plan.
00:11:08 Now the budget will be for the broken down along the regions and products. This is executed
automatically by the system via so-called data action.
00:11:17 So, let's execute it now. This is also the point where real collaborative enterprise planning starts,

00:11:27 we break down the financial numbers into detailed sales and expense numbers to create a
starting point for our domain planners.
00:11:38 For revenues, we currently have no assignment on the regional level. The system will allocate
the values
00:11:44 among the regions, based on the values available for the previous year's forecast. For the
product dimension, we already have an assignment
00:11:52 on product group level based on the values copied from the strategy. The system will further the
disaggregate
00:11:59 to the product level, referring to the previous year's forecast. We apply the same regional
distribution
00:12:06 logic for the operating expense and costs of goods sold. That means we distribute the values
among
00:12:12 the regions based on the previous year's forecast. It is special for the costs of goods sold that
the values on product level
00:12:20 are broken down based on the budget revenues to be aligned with the number of sales. Now
that the data action is finished,

7
00:12:35 let's quickly check the values generated by the system. We create a new product with a member

00:12:50 on-the-fly functionality by just using the context menu of our table. We create our new e-bike,
M600E, and we can directly assign property values.
00:13:09 We assign the mountain bike product group and make it an e-bike type.
00:13:29 Now that the new e-bike is created, we want to assign some of the budget to the new product.
00:13:35 The overall budget for the mountain bike group was already raised, within strategic planning to
reflect the necessary investment.
00:13:44 That's why we now want to redistribute the budget for the mountain bike product group to assign
some budget to our new e-bike.
00:13:52 We will leave the budget stable for our worst performing mountain bike and redistribute the
values of our better performing ones.
00:14:01 That is why we only select the better performing bikes and the new e-bike for redistribution.
00:14:09 Here we need to enter the weights for the redistribution. As our new bike will need some time to
be established in the market,
00:14:16 the revenues will not be that high in the beginning, so only a minor part of the budget will be
assigned.
00:14:25 Now we just need to apply, and the values are automatically redistributed. To make the new
budget available for all, we need to publish the data.
00:15:02 Now that the top-down budget is ready, and we have introduced a new product in alignment with
our strategy,
00:15:08 we inform our colleagues to start annual marketing planning. They will need to detail the budget
in the planning process.
00:15:18 We leave a comment at the expense overview to inform our LoB colleague about the current
status of the budgeting process.
00:15:30 Additionally, we used the collaboration panel to start a discussion with our colleague.
00:15:53 We attach the budget story so that he gets notified automatically and can directly jump into the
right place to check the budgeted numbers.
00:16:06 SAC self-service capabilities and augmented BI features set up a domain plan from scratch, in
this particular case for marketing campaign,
00:16:15 planning across the regions. For our domain plan for marketing budget
00:16:19 in 2021, we will reduce global dimensions like regions and products. Additional campaign details
will be added as local dimensions.
00:16:28 With this approach, you can leverage central structures to support efficiency and ensure data
quality across all domain plans.
00:16:35 So first, let's check out the result from the top- down budgeting process by opening the linked
story in the discussion panel and navigate
00:16:42 to the widget by clicking on the comment in the comments panel. As we can see from the top-
down budgeting, our area of marketing expenses has been an
00:16:55 assigned budget of 92.3 million. As a next step, we create our table that will serve to input
00:17:02 all our planned campaigns for 2021. Let's fast forward.
00:17:19 Together with the strategy team, we aligned to launch a new marketing campaign for our new e-
mountain bike in 2021.
00:17:27 So, we create this campaign directly here in this table. The new campaign will be created as
master data in our campaign dimension.
00:17:48 Additionally, we add the new e-mountain bike that we created during the budgeting process,
which is coming from our global dimension.
00:17:56 The budget for this campaign will be 2 million, and we just end it on the node level for 2021.
00:18:01 You can directly drill down the hierarchy to see the details across all products, or we can switch
here to see only these members.
00:18:22 Now we can share this story, including the table, with all our team members to add all planned
campaigns for 2021.

8
00:18:31 As some regions have prepared their campaign budget for 2021 in Google Sheets,
00:18:36 I will show you how to integrate this into our local domain model. Let's have a look at the
structure of the Google spreadsheet.
00:18:45 From here, we take the link to the spreadsheet, copy to clipboard, and now we can start loading
this into our domain model.
00:19:03 In the Data Management area, we connect the source of Google Sheets to our model and we
define the data mapping for it.
00:19:25 Let's fast forward a little bit here as well. In case you need to change the structure of your data,
like in this case,
00:19:38 the data format, you can do this directly during the mapping process. Finally, all our data is
mapped correctly, and we can start importing the data.
00:19:57 Now, let's have a look at the story with our loaded data. We insert a line for our total and we see
that we have a budget plan of 95 million.
00:20:17 As a next step, we will use our smart capabilities to create a comprehensive reporting.
Therefore, we start smart discovery.
00:20:25 Smart discovery will use machine learning to generate insights on top of your story. As a first
result, we see the summary of our marketing expenses,
00:20:53 and as this information is important during the bottom-up process, we will pin it to our home
screen.
00:20:59 In the time series chart below, we see which budget is required in which quarter.
00:21:05 This might be of additional value, for example, for our Treasury team. Coming back to the home
screen, we see the edit KPI of our marketing expenses.
00:21:21 Before we hand over back to the budgeting team, we want to check again with the assigned
budget from the top-down process.
00:21:28 We can directly ask the system using search to insight and using natural language search
capabilities to gather immediate information.
00:21:47 As we can see, we only have an assigned budget of 93.5 million, so we now need to align with
the global team.
00:22:01 Via the discussion panel, we get informed about an exceed bottom-up marketing budget, so let
us check the required adjustments.
00:22:14 We can see that an additional marketing budget of 3.51 million dollars is required. As we want to
hold the overall Other Operating Expense stable,
00:22:26 we lock the targeted expense value. Now we increase the marketing expenses by the necessary
adjustment,
00:22:34 and automatically, we allocate, the remaining expenses accordingly. Having allied our top-down
and bottom-up marketing expenses, we now want to switch
00:22:52 over to workforce planning to check alignment with our strategy. We have created a new model
in smart predict to better predict attrition.
00:23:04 Simply using last year's rate will not provide the insights needed to drive change in our
organization.
00:23:10 Once the model has been trained, we can review the key influencers in detail. For seniority we
see that those with five plus years' experience are most likely to leave.
00:23:19 And from a gender category we see that females are most likely to leave. Knowing that seniority
is a key influencer,
00:23:26 perhaps we'd like to allocate some money to those with five plus years' experience that are
predicted to leave based on our smart predict model.
00:23:43 We can make many different changes to our overall value driver tree, both the drivers and the
individual members, to simulate different scenarios.
00:23:52 In this case, we may look to bring our entire plant back into budget by adjusting the number of
contractors.
00:24:01 And we may, depending on the results, need to look at our yearly raise as well.
00:24:09 Once we're happy with our plan, we can run a data action, which essentially will

9
00:24:16 copy our global rates down to the employee level, and this is what's going to allow us to engage
our management team for making adjustments to their employees.
00:24:29 So here we're looking at Marco's team for the assembly operator group for pay grade six.
00:24:37 Looking at the chart, those in green are being paid more than average and those in red are
being paid less than average.
00:24:43 Here, we're going to focus on Fan Fan, who is an extraordinary employee that is paid well below
average.
00:24:50 We'll adjust Fan Fan's salary to raise to 10 percent and lower the raises for several employees
that are underperforming.
00:25:06 Here we are able to see the comments left by the planners during the planning process about
the key positions in our
00:25:12 budget, giving context to how the final numbers are derived. This immediately gives the CFO an
overview of the intention behind.
00:25:21 It's now easier for him to analyze and to decide. In this unit, you had the opportunity
00:25:28 to see how an end to end scenario for financial planning can look. It started with the overview of
the company's most important KPIs
00:25:39 and then showing a monthly forecast using predicted plan values. Furthermore, also, a multi-
year strategic
00:25:47 plan has been presented, which then was used as our basis for the annual budget. In the next
unit, my colleague, Holger, will talk about creating planning models.
00:26:01 Thank you for watching.

10
Week 1 Unit 4

00:00:06 Hello, and welcome to the session. This is unit four in week one.
00:00:11 Today's topic is about creating planning models in SAP Analytics Cloud. Data models are
actually the core foundation of each and every planning application.
00:00:21 And typically, you should spend a lot of time designing the data model and making sure that all
the business requirements can be kept.
00:00:29 From a conceptual perspective, data models in SAP Analytics Cloud follow a so-called
multidimensional modeling paradigm,
00:00:36 similar to a cube in a data warehouse environment. In SAP Analytics Cloud,
00:00:41 each data model consists of several dimensions and key figures on measures. The dimensions
are used to capture a certain aspect of the values.
00:00:50 So, for example, if you plan revenues, you want to break it down by product customer or sales
org, for example.
00:00:57 In this case, product customer and sales org would be your dimensions in your Analytics Cloud
data model.
00:01:05 Typically, you create a data model for each planning application you create in the tool.
00:01:10 So, for example, if you create a sales planning application, you typically create a dedicated sales
planning model.
00:01:16 The same is true if you create a workforce planning application, so then you have a dedicated
workforce planning, making sure that all
00:01:23 the relevant dimensions and measures or KPIs are part of this model. For traditional reasons,
00:01:31 the KPIs or measures are called accounts in the data model of Analytics Cloud. So originally, the
focus
00:01:39 of Analytics Cloud was more on financial planning and typically in finance, financial accounting,
or in financial
00:01:46 planning, you use G/L accounts as your relevant measures. And this is why the account I
mentioned was called account dimension.
00:01:54 But in general cases, you can just think about the account I mentioned as your structure where
you keep track of the different measures or KPIs.
00:02:03 Some examples for dimensions are, as mentioned, time, product, version, or location.
00:02:09 On the measures side, we typically plan on revenue, costs, sales quantities, or headcounts. In
SAP Analytics Cloud,
00:02:18 we have a dedicated modeling environment for creating the different data models. So you don't
need any external tool in order to model your planning model.
00:02:29 The models can be created in three different ways. So, first of all, you can start with a blank
model,
00:02:34 meaning you define everything from scratch in the modeling environment. The second option
allows you to create
00:02:40 a data model based on the flat file you have on your local computer. So let's say you have a
certain data sample representing the relevant structure,
00:02:48 you can create a model based on top of this data. The third option allows you basically
00:02:53 to create a model also based on a data model existing in a remote system. So, for example,
00:03:01 if you have a data warehouse system with a query or you have an S/4 system with the structures
already available, you can connect to the system and create
00:03:09 a model in Analytics Cloud based on these remote structures. The modeler is used for several
things.
00:03:16 So first of all, of course, you define the structure so the dimensions and also the KPIs or
measures,
00:03:22 but you can also maintain certain central settings like currency translation and also the definition
of the time dimension, we will come to this in a minute.
00:03:31 Typically, you can also set up security. So we will look into this topic

11
00:03:34 in a dedicated session and you can also establish semantical relationships between the different
dimensions.
00:03:40 So this is a capability we will see in a second. And in the next session, we will also look into data
imports,
00:03:48 so the data management is also done out of the modeling environment. Now, let's have a
deeper look at the concept of the dimensions.
00:03:58 As mentioned, dimensions are used to describe a certain semantical aspect of a data point. We
have the example of revenue, which is
00:04:06 broken down by sales or product and customer. Technically, dimensions or the dimension
members need to have an I.D.,
00:04:15 so a technical I.D., which is unique for each dimension member in its dimension, you can
provide a description, and then
00:04:22 optionally dimensions can also have custom properties and hierarchies. You can even create an
alternative hierarchy,
00:04:28 so let's say for reporting purposes, this might often make a lot of sense. And in these scenarios,
you can then create alternative hierarchies for dimensional.
00:04:39 In SAP Analytics Cloud, there are predefined dimension types, so as mentioned, there is this
specific
00:04:46 dimensioned type called account dimension, so this is used to keep track of the KPI structure of
your data model.
00:04:53 Then every planning model has a time dimension, you can actually have several times
dimensions, but at least you need to have one.
00:04:59 And then by default, each planning model has also a version dimension, exactly one version, to
keep
00:05:04 track of different data categories, for example, plan, budget, and forecast data. And then
optionally,
00:05:10 you can also have a dimension of type organization in order to provide a certain organizational
structure to your data model.
00:05:18 And this organization dimension comes then with certain system properties, which can be used,
for example, for currency conversion or in order to steer the process.
00:05:28 All other dimensions of type generic, and you can have basically an arbitrary number of
dimensions in your data model.
00:05:39 The members of the dimension present actually the master data, and they are typically loaded
00:05:45 from an external system, either from a centralized data governance system like a data
warehouse or from an ERP system.
00:05:55 Let me come back to these member validation combinations or to the member combination rules
in SAP Analytics Cloud.
00:06:02 As mentioned earlier, it's possible now to create semantic rules on top of dimensions. So, for
example, if you have a dimension
00:06:10 sales org and you have a dimension product and you want to sell only certain products in a
particular sales org or in a particular region.
00:06:18 So then, from a technical perspective, it doesn't make sense to allow plan data to be entered on
any combination that is possible,
00:06:26 but you want to restrict actually the combinations of product and sales org to a certain set of
valid combinations.
00:06:33 And in the modeler, you can actually specify here in the rules section these kind of member
validation rules.
00:06:39 So in this example here in the screenshot, we have the dimension sales org in entity, and
typically every sales org
00:06:47 gets an assigned entity. So each sales org belongs to a particular legal entity.
00:06:52 But of course, a legal entity can have several sales organizations. And in this case, this is the
so-called 1:N relationship.
00:07:00 And the regulation or the rule is here defined on an attribute at the sales org dimension

12
00:07:06 where every sales org then has the relevant entity as an attribute. And here in this rule definition
screen,
00:07:14 you can actually then define the valid member combinations. And the system makes sure that a
planner
00:07:19 cannot enter plan data on any other combination that is not valid. As mentioned in the beginning,

00:07:28 the account dimension has a specific meaning in Analytics Cloud, it is used to define the KPIs or
the measures of your planning model.
00:07:38 So there are, in principle, two types of measures or accounts in this case then, so you can have
the base members, which actually store the value.
00:07:46 So, for example, if you have sales as your base KPI, so then you store just the values for this
KPI.
00:07:55 But sometimes you also want to calculate based on these base KPIs, and then you can create
here calculated
00:08:01 measures or calculated accounts as part of the account I mentioned. And in this case,
00:08:06 the system will calculate the value of these calculated accounts on the fly based on other KPIs
that are defined as base members here in the data model.
00:08:17 A very important aspect is that the values of these calculated KPIs are determined at runtime
and the results
00:08:24 of these KPIs are not persisted in the database. So this is the difference to the base KPIs
00:08:29 or the base accounts defined here in the account I mentioned. The formula or the rule
00:08:36 for the calculation is specified here in the formula attribute of the account members, and you
have also here access to more complex calculation logic.
00:08:43 So we have to look up formulas to read certain values. You can also provide here conditional
statements so you can
00:08:51 specify certain complexity here in the account I mentioned for the calculated members.
00:08:59 As I mentioned in the beginning, you can do certain central settings for the model here directly in
the modeling environment,
00:09:05 so these settings relate to data access and privacy. So we have covered this in a dedicated
session.
00:09:11 But you can also specify certain attributes of the time dimension, for example, if you use fiscal
years
00:09:16 instead of calendar years, so you can specify then the offset, for example, if your fiscal year
starts on July 1st instead of January 1st,
00:09:23 so then you can specify this here as an attribute of the time dimension. Another general setting
here is related
00:09:30 to the currency handling or currency translation. And this is where we have a detailed look.
00:09:37 The concept of currency in SAP Analytics Cloud is attached to the system property,
00:09:44 to a currency property, which you can switch on for a particular dimension. So let's say you have
sales org as your
00:09:51 organization or dimensioning your data model. And of course, each sales org has a particular
local
00:09:57 currency, and you want to allow people to plan in their local currency. For example, you want to
have the Swiss sales org to plan in Swiss francs,
00:10:05 whereas the US American sales org will plan in US dollars. And in this case, you can switch
00:10:11 on the currency attribute for the dimension sales org, and then this has an impact on how the
system actually interprets
00:10:19 and stores the number in the database. We will come to this on the next slide.
00:10:24 The second important setting then, besides this currency attribute, is related to the rate table.
00:10:29 So in the system you can maintain actually several rate tables, storing the currency translation
rates.

13
00:10:35 The rates themselves can be time dependent and version dependent, so you can have different
actual rates from budget.
00:10:42 Or you can, of course, then also have different rates throughout the year. And the third important
setting here is the date dimension.
00:10:50 Of course, currency translation always has relation to the time at which you want to translate the
values.
00:10:57 And since the data model can have several time dimensions, you need to specify the time
dimension
00:11:02 that is relevant for the currency translation. And then the fourth parameter or the last
00:11:09 parameter you can do here in the settings is related to the default currency. Since the system
always translates at runtime from the local currencies
00:11:18 into reporting currency, you can actually maintain your default currency. And this is then used
later on in the report to translate the local
00:11:25 currency into this default currency or into this unique reporting currency. So, now let's have a
look at how this is used
00:11:36 in the reports or the planner in the front end. So, we have here in the screenshot our sales org
dimension and sales org here
00:11:44 is used as the relevant dimension for the currency translation, you see that we have switched on
here the currency attribute,
00:11:51 which is appearing here as a fourth column here in the screenshot. And each sales org, of
course,
00:11:58 has its local currency, for example, here the sales org SO001 has U.S. dollar, whereas sales org

00:12:04 SO002 has euro as the local currency. And this has an impact now as the data is stored in the
database.
00:12:13 So each value now or each data point is interpreted as local currency and the local currency is
actually coming
00:12:22 from the attribute of this sales org dimension in the database. And now if you want to report
00:12:29 in the story, we need to specify which is our reporting currency. So by default, this is the default
currency, which is set as a model parameter.
00:12:37 So the system will translate the values on the fly from the local currencies, which are stored on
the database, virtually into this reporting currency,
00:12:46 which is the default currency, but can be any other currency as well. This is then up to the user
in which currency he wants to see
00:12:54 the report or in which currency he wants to enter his plan data. And with this, I want to stop my
presentation
00:13:01 and jump directly into the tool so that you can see also the features I talked about here in action
in an SAP Analytics Cloud system.
00:13:09 So, in order to create a new data model, I navigate to the main menu and then I select Create
Model.
00:13:18 As discussed, there are three different workflows, how you can create a model, and I will start
here with a blank model and create
00:13:24 everything from scratch directly here in SAP Analytics Cloud. We see here now the modelling
environment initially, of course,
00:13:33 we just have a version dimension and by default the time dimension, since we always need to
have one time dimension, at least in our data.
00:13:42 Now, as the next step, I'm going to create additional dimensions here to my data model. I
navigate to the toolbar, into the Edit section, and then I can actually either
00:13:52 create a new dimension or I can also add an existing dimension, and in this case, I would reuse
existing master data in the system.
00:14:01 So when I create a new dimension, I can basically select whether it's public or private, and
public dimensions can then be reused later on.
00:14:10 Here in our case, I'm going to create everything from scratch, so I create here a new dimension.

14
00:14:17 I need to provide here a name for the dimension. I call this PRODUCT, since I want to create
here a data model for sales planning.
00:14:23 So I'm going to create a sales org, product, and customer dimension. And then later on, I create
KPIs in order
00:14:30 to plan the cross sales and discounts and stuff like this. So, here I select the dimension type, so
in my case, the PRODUCT dimension is
00:14:41 just a generic dimensioned, so no specific attributes or anything attached to it. And here I can
select whether this dimension should be public or private.
00:14:50 As mentioned in the beginning, the public dimension can then also be reused in other models.
00:14:55 And this way I can basically share master data between different planning models. In our case,
the PRODUCT dimension
00:15:01 remains private because I do not want to reuse it later on in any other model. Once the
dimensioned is created,
00:15:10 I can select here in the Dimension Settings additional properties, or I can do some settings, so
for example, I can provide here a description
00:15:20 and then I can also create additional properties or hierarchies. And in our case,
00:15:25 since the product dimension is structured in a hierarchical way, I want to create a hierarchy from
my PRODUCT dimension.
00:15:32 In general, there are two different types of hierarchies and SAP Analytics Cloud, so there's the
so-called parent-child hierarchy
00:15:39 and then there's also a level-based hierarchy. In this case, I select here parent-child hierarchy,
and with the parent-child hierarchy
00:15:46 for each member, we basically specify its parent member. And then the system can this way
00:15:52 implicitly create the hierarchy from this definition. I give this a technical name and then the
description I just call this here hierarchy.
00:16:11 Now I can open the dimensional details. And here I will see then my dimension members.
00:16:18 Initially, of course, there are no members maintained, I can now populate this dimension with a
corresponding master data.
00:16:25 In a real system or in a real scenario, of course, normally you load this master data from a
central system.
00:16:31 So this is actually also something you will see then in the next unit. But for now, I just copy over
here from the local flat file.
00:16:38 So I navigate you into my Excel spreadsheet there I prepared the master data and just copy over

00:16:45 my PRODUCT dimension master data. So just do copy paste

00:16:56 and then I have populated my dimension here with the individual product IDs. We can also now
have a look at the hierarchical structure,
00:17:04 so I can navigate here through the hierarchy view. You will see that we have here the total load
at the top,
00:17:10 then I have a different product groups, so I have cross bikes, cruise bikes, mountain bikes, and
so on.
00:17:15 And then in the different product groups, I can further drill down until I come here to the lowest
level,
00:17:20 which is representing then my individual SKUs. So now let's continue.
00:17:28 I'm creating my additional dimension, so I select here again new dimension, then I want to have
besides product,
00:17:36 I also need to create a dimension for my customers. So I call this CUSTOMER,
00:17:44 I make it a generic dimension again, and again, I can select whether this should be public or
private.
00:17:50 I choose here private again, so I leave this empty. For the CUSTOMER dimension I do the
same.
00:17:56 I actually create here a description and this time I want to create a level-based hierarchy and a
level-based hierarchy is based on its attribute,

15
00:18:06 and the attributes represent basically the level. So here for our customer dimension, I want to
have attributes specifying
00:18:13 whether the customer is a retail store, or is a wholesaler, and whether it's then a large retailer or
medium retailer.
00:18:22 And based on these attributes, I create then my hierarchy. So first of all, I need to specify here
my custom
00:18:28 properties before I can create this level-based hierarchy. So, I just create here three attributes
00:18:37 to represent the three levels of my hierarchy. L1, L2, and finally, L3.
00:18:51 So, again, I navigate here to the dimension details and then I copy over my customer master
data.
00:19:03 From my Excel spreadsheet. Just copy paste here into these dimension settings.
00:19:20 Now we see we have here populated also these custom properties, and now I'm going to create
a
00:19:26 level-based hierarchy this time based on these three different attributes. In order to do so, I go
here to the Dimension Settings here on the right,
00:19:34 I click Create Hierarchy, and then I choose level- based hierarchy this time. I create a new
hierarchy.
00:19:42 I provide a name and then, at the lowest level of the level-based hierarchy, there are always the
member IDs, and then I
00:19:49 create the additional levels on top using these custom properties. And you see now at the lowest
level,
00:20:00 we distinguish between medium retailers, large retailers, and wholesalers. And then on level
two, we distinguish between direct and indirect channel.
00:20:07 And then at the top level, I have just one member, so one distinct attribute value, representing
now my total node.
00:20:20 So, then I create now our SALESORG dimension.
00:20:37 Again I pick generic and just keep this as a private dimension. I provide here in dimension
settings a description.
00:20:49 Of course, I want to have a hierarchy also for reporting purposes. And then I want to enable
currency also
00:21:00 because every sales org should be able to plan in its local currency afterwards. So this is why I
toggle on the currency property.
00:21:07 This will now add here an additional system property to my SALESORG dimension.
00:21:14 I also add another custom property. Because I want to link my SALESORG
00:21:21 to the particular legal entity, because maybe later on we want to translate or transfer the sales
plan into a financial plan,
00:21:28 and typically in a financial planning model, you have the legal entity as the organizational
dimension rather than the sales.
00:21:36 So I navigate here to the dimensioned members, and you see, initially it's empty, but we have
now additional property,
00:21:43 so we have the system property for the currency code and then we have the custom property to
specify the legal entity for each sales org.
00:21:54 Again, I'm switching now to my SalesOrg tab here in the spreadsheet. I copy all of the values.
00:22:05 And then I paste it here, into my SAC model environment. So, as the next dimension, I'm going
to create this legal entity so
00:22:15 that we also have it here and we can immediately derive it from the sales data. And this time, I'm
going to create here a dimensional type of organization.
00:22:42 I provide here a description. Again, I want to have a hierarchy.
00:22:51 Also to be able to report the different regions. And then I create another property, this time
COUNTRY.
00:23:03 Also later on to build my report. So I navigate here again into the dimension members.

16
00:23:16 And then I copy over from my spreadsheet. So, at this point in time, we basically have a model
with a version dimension,
00:23:32 a date dimension, and then we have PRODUCT, CUSTOMER, SALESORG, and ENTITY. And
what is missing now are our business KPIs we want to actually plan on,
00:23:40 so for sales, for example, but also quantities, prices, and discounts. And for this, I create another
dimension
00:23:50 called the account dimension. So here I click add new dimension
00:23:57 And this time I choose as the dimension type, I choose type account, and here I call this
MEASURE because this is not a financial planning application, but
00:24:07 more a self-planning application from a terminology perspective. Typically, people use more the
measure term instead of the account.
00:24:17 So I provide here a description again, and then before I add the KPIs here
00:24:23 into my account dimension, I want to switch on my currency conversion. So, I navigate here to
the model preferences
00:24:32 and then in the model preferences. I navigate to the tab for the currency settings.
00:24:41 I toggle on the currency translation, you see here that I can specify default currency, so this is
then used
00:24:48 in the report to translate from the local currency into a reporting currency if nothing else is
specified.
00:24:54 Then we pick SALESORG as the dimension with the currency property, since we have two
dimensions
00:25:00 I could also go for ENTITY, but in sales planning, the sales organization is the leading
dimension.
00:25:07 And I can pick here a rates table so you can see I can have your money in my system. So I pick
this one.
00:25:14 And since we just have one time dimension, I pick this dimension as the relevant one for the
currency translation.
00:25:25 So now I'm opening up here my account dimension, and I can now specify my KPIs. So again,
I've also prepared this, so I'm going now to copy over from Excel,
00:25:48 and paste here into my account dimension. So, we have here four different KPIs, so we can
00:25:55 plan on quantities, prices, then derived the gross sales, and then also plan the discounts in order
to get to the net sales later on.
00:26:05 What we can see here is that some of these accounts are actually currency related.
00:26:11 So if we scroll here to the right, we see that, of course, DISCOUNTS, GROSS_SALES, and
PRICE
00:26:16 are related to currencies, whereas QUANTITY, of course, does not have any currency.
00:26:21 This way we can actually then control which of my KPIs are later on translated using the
currency conversion.
00:26:30 Now, I can navigate back here to the overview page. At this point in time, we have now defined
the structure of our data model.
00:26:38 So we have here by default the version dimension, we have a date dimension.
00:26:42 We have defined our KPIs in the account dimension. Then we have PRODUCT, CUSTOMER,
and SALESORG as generic dimensions,
00:26:48 and we use here ENTITY as the organization dimension. So at this point in time, we have the
structure,
00:26:54 we have also the master data uploaded here just by copy and paste. And I can save my data
model now.
00:27:03 So I just give it a name, so let's call it just SALES_PLANNING, I can provide here a description,

00:27:11 and then I can save it here into the file repository of Analytics Cloud. As a next step, I'm going to
create these member validation rules.
00:27:24 In order to do so, we first need to switch this on. For this, I navigate here to the Model
Preferences.

17
00:27:35 And then I go to Access and Privacy, and then I will switch on to this toggle for the validation
rules.
00:27:45 I need to save this again. And then this top strip Rules should be
00:27:53 enabled and I navigate now into the design time for these member validation rules. I can create
here a new rule
00:28:04 I call this Sales Org / Entity Mapping. And what I want to establish now here is a combination
rule between SALESORG
00:28:18 and legal entity, because every sales org is assigned to exactly one legal entity. So it doesn't
make sense to plan on any technical possible combination,
00:28:29 but for sales org and legal entity only certain combinations are valid. So, I add here my
SALESORG dimension.
00:28:40 And this is where now the attribute entity is used, because via the attribute we actually establish
the link into the entity dimension.
00:28:50 And I picked this attribute, so exactly this Entity attribute we created just before,
00:28:55 and this should match then to the dimension members or the dimension IDs or member IDs of
my ENTITY dimension.
00:29:07 And you can see now the valid combinations, so Sales Org US is only allowed to plan on the
combination sales of US entity 001.
00:29:18 So I can save this rule. And now we have the data model in place
00:29:27 and as a next step then in the next unit we will see how we can populate also transactional data
into this model.
00:29:34 So stay tuned and I hope to see you next time. Thank you and goodbye.

18
Week 1 Unit 5

00:00:05 Hello, and welcome to the session. This is unit five of week one,
00:00:09 and the topic we want to cover today is data integration and connectivity. So, this is an important
topic for planning because typically a planning solution is
00:00:18 integrated into the overall IT system landscape. And since SAP Analytics Cloud
00:00:23 is a pure cloud-based solution, we need to think about the different source systems which fit the
data into the planning process.
00:00:30 So typically, master data is coming from a central system, so either a data warehouse or an ERP

00:00:35 system or from any other central database where you maintain the master data. But also
typically in planning you need
00:00:42 some transactional data as a baseline for your plans. And this data is typically then also coming
from another system,
00:00:48 so from a transactional database or from any other IT system in your landscape. Depending on
the system type,
00:00:55 there are different ways to set it up. So on the one hand, you might integrate also with other
cloud systems.
00:01:01 Think about an S/4HANA Cloud system or a SuccessFactors system, which is also running in
the cloud.
00:01:06 And in this case, of course, the integration is a lot easier because here we are talking about a
cloud to cloud integration.
00:01:15 If the system is sitting behind the firewall, so it's an on-premise system in the customer's local
network, then,
00:01:20 of course, the system set up gets a little bit more complicated. In this case, we actually need to
bridge between the public Internet and the local
00:01:28 corporate network, which we do with a component called the cloud connector. So this
component has the functionality to bridge between
00:01:36 the two network segments and to allow Analytics Cloud to import data from the source system
running behind the corporate firewall.
00:01:45 Then it depends also a little bit on the system protocol, which is used to extract the data.
00:01:51 So some systems use proprietary protocols, others use HTTP-based protocols, and then,
depending on these technologies, you might also need to install here
00:02:01 a so-called cloud agent, which is an additional component that needs to be deployed inside the
customer network.
00:02:07 What is also important to understand is that in planning, we always rely on acquired data.
00:02:13 So from BI, you might know that there's also a so-called live connectivity, which allows to
consume the data directly
00:02:19 from the source system without any data replication, in planning we actually rely on the
acquisition of the source system data,
00:02:27 meaning we need to import the data into our Analytics Cloud data model. The data management
is actually then done
00:02:37 inside the modeling environment, so there's a dedicated tab called Data Management and in this
Data Management, you can set up different import jobs.
00:02:46 So, for example, for both master data as well as transactional data. You can also connect
several systems.
00:02:52 So, let's say if you're in a sales planning scenario and you have local customer databases in the
different regions, you can then also connect to several
00:03:00 systems in order to populate your customer dimension. The different master data jobs can also
be set into relation,
00:03:07 so for example, if you have several jobs to populate a dimension, so let's say there is an API to
fetch

19
00:03:13 the master data and then there is another API to fetch the hierarchy information, then of course
you want to cancel the whole sequence
00:03:22 if one of these import jobs fails, and these kind of dependencies, you can actually also set up
when you define the import jobs.
00:03:29 What you can also do is that you schedule the import jobs on the regular cadence, so let's say if
you have frequent master data
00:03:36 updates that you want to consider in your planning tool, so then you can also schedule here the
import jobs to run, let's say, every day or every week.
00:03:46 So this is also possible here in the data management view. Before the data is actually stored
inside the tables of the database
00:03:59 for the SAC models, you can actually define so- called data wrangling. So this means that you
can define transformations and define how the source
00:04:09 data is actually transformed because it goes into the data model. So what we see here on the
screenshot is
00:04:16 a data wrangling environment for a particular import job. On the left-hand side, you see the fields
that are exposed
00:04:22 from the source system, and then in the middle screen for these tiles, we see actually then the
structure of our SAC planning model.
00:04:30 And, of course, you need to map the fields of the source system API here to the different
dimensions of your data model.
00:04:36 You can then define also additional import settings and you can also define data
transformations.
00:04:42 So let's say if you want to have a particular search and replace algorithm so that you want to
map certain IDs
00:04:48 to other IDs, so then you can define these in these transformational wars and the transformation
rules are also stored as part of the import job
00:04:56 and are rerun whenever you import new data from the source system into your SAP Analytics
Cloud data model.
00:05:05 What is specific for planning is that you can also export data. So this is typically not the case in
BI
00:05:12 or analytic scenario, where you just import data from a source system and then you create your
stories or your analytical application.
00:05:18 But since in planning you typically generate new data or you change existing data,
00:05:23 sometimes the data is also needed in other systems for, let's say, subsequent processes or for
any other business process.
00:05:32 And this is why in planning that is also often the requirement that you are able to retract the plan
data back into the source system.
00:05:40 And this is also what is possible via the Data Management environment.
00:05:44 So there is a section for the export where you can actually then push the data back into the
source system, for example, into an S/4 system.
00:05:52 And the process here looks very similar to the import side. So there's also a so-called wrangling
or mapping step where you define
00:06:01 the source or the mapping between the source structure. So in this case then, the data model
inside
00:06:06 SAP Analytics Cloud to the structure in this remote system, so let's say a particular table in the
S/4HANA system.
00:06:13 So you can also apply certain filter settings so that you just copy back your budget information,
for example,
00:06:19 and not the whole data that is sitting in the SAP Analytics Cloud data model. And as a next step,

00:06:25 I want to also show you this in the real system, not just with theoretical slides. So, now I'm back
in the data modeling environment where we left end of last unit.
00:06:36 And at this point in time, we have here the structure defined. So we have a data model with
different dimensions.

20
00:06:41 We have also loadmaster data. Just via copy and paste from a workbook.
00:06:46 And now as a next step, I want to populate this data model with transactional data. For this, I
navigate to the Data Management view.
00:06:55 And in the Data Management view, we can actually then import create jobs here for the
transaction data.
00:07:01 You see the section below where we can also then export data later on. But for now, we just
want to import the transaction data.
00:07:09 When I click on this button, I see two options. I can either upload here from a file or I can
connect to a system,
00:07:15 so this is here the item data source, and load from an external system, so, for example, from a
data warehouse system or from an ERP system.
00:07:23 Here in this first example of our sales planning model, I have prepared here a local file and I'm
going to upload the data here from this local file.
00:07:34 I navigate to Select Source File, and then I click here my prepared Excel workbook, since this
workbook can have several tabs,
00:07:43 I choose the one that contains the Transactional Data. And then I hit the Import button.
00:07:53 What is happening now is that the system is actually uploading the data into a special region of
the system,
00:07:59 so here into the Draft Sources, and from the Draft Sources, I can then create my transformation
rules and the mapping between the columns
00:08:08 of this workbook and the dimensions of my data. Once this is uploaded, I can open up now the
data wrangling environment,
00:08:21 and in the data wrangling environment, we can define the mapping between the source file and
the dimensions of our data model.
00:08:34 So initially, the system tries to map the fields or the columns here of our source file to the
individual dimensions of our data model.
00:08:43 So normally, if the naming convention is similar then this should also be successful.
00:08:49 If this is not the case, you can actually undo the mapping and establish this manually.
00:08:54 So, for example, let's assume this product dimension was now mapped in the wrong way. So
then I can just hit here on this icon,
00:09:04 and then see the mapping is released and I can now just drag and drop the product here from
the left-hand side into this middle screen.
00:09:16 For the import, there are different global settings, so, for example, you can specify how the data
is treated,
00:09:24 whether the existing data should be updated or whether you want to clean and replace existing
data with the import data.
00:09:30 So these are all kind of different settings you can specify.
00:09:33 What we need to do here is we need to map the version. So I click on the Version dimension.
00:09:39 Since we just have actual data here in our local file, I just need to map this version.
00:09:44 But in general, there could be, of course, several versions in my source file. And then we can
also have a preview, so I
00:09:51 can switch here the layout between this card view and the grid layout. So here we see then a
preview of the data to be imported.
00:09:58 So we see how the Excel file structure. So basically, for each
00:10:03 measure we have here the data record for the different dimensions. And then finally, we have
the numerical value then for gross sales, in this example.
00:10:16 Once this is done, I can click the Finish Mapping, but in case I want to do some transformations,

00:10:25 so for example, I want to remap certain product IDs, I can specify here on transformation.
00:10:31 So, for example, I want to have here. A replacement of particular product IDs,
00:10:38 so I could specify here then this transformation rule. But in our case, I leave it basically as is,
and then finalize the mapping.

21
00:10:46 And now the system actually moves data from the file into the underlying database table of our
data model.
00:11:01 I get here final confirmation and then the data is loaded.
00:11:20 So I get then summary, so in this case, we imported 224,000 rows into our data model,
00:11:26 and I also get in here the summary of this import job. This was here an ad hoc import, normally,

00:11:34 if we import data, let's say, from a central IT system, so from an ERP system, of course, you
want to schedule normally the import
00:11:41 so that you rerun these import jobs on a regular basis. So now let's have a look into, let's say, a
more realistic example.
00:12:09 And in this scenario, I've created a financial planning model, so we see here that we have
00:12:16 the account dimension called G/L Account. So this is representing then my G/L Account from the
general ledger.
00:12:21 We have a Date dimension, then we have a Cost Center, Functional Area, Profit Center, Trading
Partner, and we have a Company Code.
00:12:29 And in this scenario, I want to actually load first master data from an S/4 system and then later
on
00:12:34 also the transactional data as a reference for my planning application. So in this case, I navigate
then here to our organization dimensions.
00:12:43 So to our Company Code dimension. You see, initially this is all empty,
00:12:48 so there is no master data loaded so far. And then I navigate here again to the data
management view.
00:12:58 And now I want to import my company code information from my central ERP system, in S/4
system in this case.
00:13:05 So again, I go here to the import icon and then I click Datasource, and this time I click here the
S/4HANA Connector.
00:13:14 Of course, someone or the administrator needs to create the connection first, so this was
already done in the step before.
00:13:27 And then the system is actually pulling the meta information from the S/4HANA, so basically the
catalog of existing OData services that are exposed,
00:13:35 and then I need to pick the right one for the company code information. So we can have a look
at the different APIs that are exposed,
00:13:46 and here I already see the one for the company code, which I can select then.
00:13:53 And since this API exposes also a lot of fields, which I do not need here in my application,
00:13:59 I can select the ones that I want to take over. So here in our case, I want to have the company
code.
00:14:05 So this is going that into the ID of my dimension members. I want to have the company name as
the description.
00:14:12 So then we have the country key as an additional attribute. So maybe for reporting purposes.
00:14:18 And then, of course, I want to have the currency information describing the local currency of the
particular company code.
00:14:35 In principle, this is the same workflow as we have seen for the local file import. So now I'm
getting here into this data wrangling environment.
00:14:43 I see a preview of the uploaded or the extracted data. So now this information is coming from the
S/4 system.
00:14:52 We see the different IDs for the company code, we see then the name of the company, we see
the country key,
00:14:58 and we see then also the currency. So these are the fields we have actually picked.
00:15:02 And now here on the right-hand side, I need to map those fields here to my dimension
properties.
00:15:08 So here the company should go to my dimension ID, and then in the next step, the company
name should go to the description,

22
00:15:17 the currency and the country key have been mapped already. And then I click on Finish.
00:15:34 And then the system is loading the data, and once this is done, I get to the success message
and in this case, of course, only 23 rows
00:15:41 have been imported here into the dimension of our planning model. If I wanted to now set up a
regular import, so I can set up here and basically
00:15:53 schedule, so you can see that I want to have here maybe the repeating frequency. You can then
set up the cadence, you can specify additional properties,
00:16:04 and then I would have regular import job running them on the cadence I have specified. But I
leave this for now.
00:16:13 The workflow for transactional data is actually exactly the same. So let's have a look here again.

00:17:46 So, the workflow for importing the transactional data is actually quite similar.
00:17:51 So now let's take a look. So our Company Code is now properly imported,
00:17:58 so we have 24 records and now I want to import the transactional data. So here I need to be
then in the model overview, not on a particular dimension.
00:18:07 And then if I navigate here to Data Management, we can actually now set up an import job for
the transactional data.
00:18:15 So I navigate again here into the import menu and then I pick Datasource and again I go to the
S/4HANA connector.
00:18:26 I pick my connection, and then I will see the catalogue of the available OData APIs.
00:18:36 And this time I'm interested in the journal entry, so this is going to the universal journal of the
S/4HANA system.
00:18:44 And I pick here this particular OData service. And now the system is pulling all the fields from
this OData service,
00:18:54 and if you scroll down, you see that these are a lot, probably most of the fields are not relevant
here for my planning scenario.
00:19:02 So I start here with the measures coming from the S/4 system. So I want to have, of course, the
amounts in company code currency.
00:19:14 And then I need to find, of course, my company code. We have cost center.
00:19:26 We have the fiscal year period, so this is the time dimension. We have the G/L account, we have
functional area.
00:19:39 You need to specify the ledger, so this is going into the filter. And then I think we had profit
center
00:19:53 and the trading partner. So, the ledger, I'm going to filter, so this is
00:20:02 coming from the standard ledger, so 0L. And we can, of course, also specify
00:20:15 the time dimension, so here I need to go back to our fiscal year period and say this should be
greater than,
00:20:25 let's say, 001/2018. And less than 012/2018.
00:20:41 So then I can hit the Create button. And now the same workflow is executed.
00:20:49 So first the data is loaded into this, let's say, staging area, into the draft sources, and then from
there, I can define my data wrangler.
00:21:11 So I can map then here the fields that have not been automatically mapped, so this looks good
now, the system was able to map all
00:21:21 the fields of my OData service to the relevant dimensions of my data model. Here we can have
the same
00:21:28 import settings as we've seen in the flood file upload. And, of course, we are importing here
directly into the actual version.
00:21:37 So I hit Finish Mapping. And now the transactional data is
00:21:41 then loaded from the S/4HANA system into our SAC planning model. And once this is done, we
get here the summary of the import job
00:21:56 so that it was successful and now we have populated our planning model with transactional data
from an S/4HANA system.

23
00:22:04 In a real scenario, of course, I can also set up here a regular cadence for this import job and
then on a constant basis,
00:22:11 the data would be pulled from S/4 and transferred here into my SAC planning. If I want to export
then later on the data,
00:22:19 so let's say we have run our planning cycle, then actually the same workflow will apply.
00:22:24 So I come here to the data management and then I can also set up here an export job to see
that we support here exports to different system types.
00:22:33 So you can export to BPC, you can export to SAP IBP, and S/4. And we have also now export
capability back into SAP BW.
00:22:44 So this is behind this export OData services. And with this, I want to finish this session.
00:22:52 Thanks for attending, and hope to see you next time.

24
Week 1 Unit 6

00:00:05 Hi, my name is Derek Johnson, I'm a product manager for SAP Analytics Cloud.
00:00:09 In the session, we're going to be talking about Ad-Hoc Planning and Version Management.
00:00:13 First thing I'd like to focus on is data entry. And we're going to have a number of exercises as
part of this openSAP course
00:00:20 where you're going to be asked to do some updates. So feel free to kind of extend what you're
doing by trying different approaches,
00:00:28 such as adding a plus and minus a percent, plus or minus values. You can also trade key words
like, 100M for one hundred million.
00:00:37 If you'd like to, you can also enter in values at parent nodes as well as child nodes as well,
00:00:42 and see how the system disaggregates or aggregates your data accordingly. So have fun.
00:00:49 I would encourage you to really take some time and extend what you're doing beyond the basic
exercise and explore the system
00:00:58 by just simply testing out some different data entry strategies here as well. One of the things
that's changed in the
00:01:11 in the last few months is the addition of the new planning panel. So what we've done is we've
replaced
00:01:17 the Assign, Distribute, and Spreading functions that we previously had. They were not really well
understood by a lot of our planning users.
00:01:27 And this has now been replaced by the planning panel. And this is a much more intuitive way to
make complex changes from a planning
00:01:36 perspective in a way that's very intuitive for a planner. And we're going to be looking at this as
well, and there's a lot of capabilities
00:01:46 in terms of how we can select our sources, our targets and our update methods that will likely
extend beyond all of our exercises.
00:01:55 So I would encourage you to run through the exercises on the planning panel, but feel free to
take some time and explore that as well.
00:02:07 One of the things that's very useful if you are exploring is to be able to compare different
versions.
00:02:13 So I know that when I am just exploring the system, a lot of times what I'll do is I'll create a
private version and then I'll be making
00:02:20 my planning changes just to test how the changes will work with the data. And using the table
calculations,
00:02:30 we can go right in here and we can create some simple variance calculations to be able to better
understand the changes that we've made.
00:02:36 And I would encourage you to do that as well. But know that there are other table functions
00:02:41 that we can use and you should consider exploring as well. And the options really depend
00:02:47 on how many columns in this case are selected. If two columns are selected, you'll have a
certain set of options.
00:02:53 If you have more than two columns, you'll have additional options that are possible as well.
00:02:58 So feel free, take some time, explore these capabilities as well, and we can extend our learning
well beyond our exercises.
00:03:10 I'd really like to take some time and focus on version management, so from an ad-hoc planning
perspective,
00:03:15 this is really one of the key areas. So we have two basic types of versions.
00:03:19 We have a public version and then a private version. A public version is available to anyone
00:03:24 that has access to the model version and that slice of data. And we have the private version,
00:03:30 which is only available to the planner who created it. So it's like a personal sandbox.
00:03:37 Now, the nice thing is with private versions, we can choose to share them with our colleagues as
well.

25
00:03:43 So just know that if you're thinking about private versions, the two basic use cases in this case
are
00:03:51 just for your personal exploration and planning activities, and you can do a scenario planning in
a private version as well.
00:03:58 But you can also share them with colleagues as well if you're working on more complex
scenarios or troubleshooting some business issue.
00:04:08 And really, you can now, if you do create private versions, and we'll be doing this plenty of times
through our exercises,
00:04:14 we can always convert those into a public version as needed through the publishing process.
00:04:21 And we'll be taking a look at that throughout the exercises. Now, I know the session is on ad-hoc
planning, so I'd like to just take a minute
00:04:30 to also discuss the version management and the history function. So from a public version if you
decide to make some updates
00:04:37 that are extending beyond what's in the exercise, that's okay. If you need to revert that back,
00:04:44 you simply have to go into the version management history and then we can step back through
those functions to go back to a previous change.
00:04:51 So for a public version, that works well until we publish the version, in which case that's the last
step we can go back to.
00:04:58 And private versions, we have a longer history list to which to revert back our changes.
00:05:04 So feel free to explore the system in really any way possible just to get a better understanding of
how
00:05:14 our planning solution works and know that you can always go into version management and the
history function and start to undo some of your changes as well.
00:05:23 So have fun and really take advantage of your time of going through the exercises to really
understand what our capabilities are.
00:05:35 The last thing I just wanted to make note of is just that as an administrator, if you need to delete
all your private
00:05:42 versions that you have for a model, it's very easy. You can do that through the Model
Preferences here as well.
00:05:49 So this is available for you as a user that's managing your activities,
00:05:55 but also it's available as a system-wide capability as well and can be useful for managing private
versions in large, complex environments.
00:06:08 With that I'd like to go into the demo and we're going to focus on our version management
because I think that's the piece often neglected in our exercises
00:06:17 and I think it's very powerful in terms of being able to kind of revert back the state. So if you are
playing around and having fun exploring your system,
00:06:25 you can always revert back your plan to a specific state that matches the exercise. With that,
let's go to the demo.
00:06:34 So in this session, we're going to be looking at how to do some basic ad-hoc planning within
SAP Analytics Cloud
00:06:41 and you'll have an exercise that goes through a lot of the same core concepts we're going to be
discussing here, it will go through them in greater detail.
00:06:49 So the big thing I want to demonstrate as part of this demo here is how to use version
management.
00:06:57 Inevitably, you'll be going through an exercise and you'll realize you've entered the wrong value
or something just doesn't look right and you want to undo
00:07:06 the step that you've taken as part of your planning activity. So we're going to show you how to
take
00:07:12 advantage of version management to undo some of those changes. And as part of that,
00:07:16 we're going to cover how to make changes to plan values directly in a table. And we'll cover the
planning panel as well
00:07:23 and how we can do some of the updates there. And we'll look at how we can use version

26
00:07:28 management if we ever want to revert any of the changes that we've made. Over here we see
our sales forecast, maybe we want to use the
00:07:42 new planning panel, we can do that here as well. Let's just say we want to distribute our
00:07:48 values for our cruise for Q4 and maybe we want to distribute that value to the other members at
the same level.
00:07:58 So we're just going to make some changes here that just to keep this pretty straightforward, do it
by input weights,
00:08:05 just to make this simple we'll do one, two, three, and four. So now we get our relative weights,
we have our value we're distributing.
00:08:15 We see that we're going to be appending this value versus overwriting it. So if we overwrote that
value, the total would be,
00:08:24 in this case, the 37 million we're going to choose to append it just so it adds to the existing
values here.
00:08:31 So we can just select the apply. And here you can see that the value has
00:08:40 now been redistributed from the cruise line to the other lines here as well. And if we want to
make some additional changes, easy enough to do as well.
00:08:49 So we're going to just do plus 10 percent as mentioned, we can do plus, you know, minus by
percentage.
00:08:55 We could also add to the value directly in here as well. If we wanted to add, let's say, 30 million,
we could do that.
00:09:02 If we wanted to use, you know, put things in thousands, we could enter 30 K.
00:09:10 Likewise, if something was in the billions, we could put, in this case, 30 B, it would have been 30
billion.
00:09:15 So we're just working in millions here, which would be 30 M, but we just take the default when it
shows
00:09:21 up, which is includes some billions in there. So let's just say we've done a number
00:09:27 of these planning changes and we would like to undo some of the changes that we've made,
and perhaps it's just that we've keyed
00:09:35 in the number very poorly and we kind of like to reset things. So we can do that through version
management.
00:09:41 And so here's our public and our private versions. And here we're just going to go into the menu.

00:09:48 And we can take a look at the changes that we've made. So you can see that
00:09:54 it gives a list of some of the changes that we've made as part of our planning process, and we
can just kind of step through these
00:10:03 these changes and kind of revert some of them, as we would like. And we're just going to do that
now.
00:10:10 So you see, as we do this, the values will show up in yellow just to kind of indicate where things
are changing in terms of our value.
00:10:20 We could have also just reverted back to the starting place, we select Revert. It's just going to let
us know that it's
00:10:26 going to take us back to the last time we published it. And you can see this is really our starting
point here for this version.
00:10:36 So the same thing works for private versions as well, and we'll take a look at that. Here you have
the Undo and you have the History.
00:10:45 You could step through these changes here as well in a very similar fashion. So if you are
exploring planning
00:10:53 and you're worried that your version of the values will start to differ significantly from what's
shown in the exercise, or you'd like to try some
00:11:01 different scenarios, really I encourage you to do so. And, you know, have fun.
00:11:07 If you would like to revert those changes back, you can easily do that with our version
management.

27
00:11:15 While we're here, some other additional things you can do is if you've created any additional
table formulas, those are easily visible.
00:11:27 You just have to select the formula here in hindsight to review the definition. You'll have your
formula showing up in the menu.
00:11:37 So by selecting that, we'll have this menu bar and we can look at the definition here as well. So,
if you're creating additional formulas, or table
00:11:48 calculations, and you'd like to review the definition, you can do that at any time. And sometimes
that's useful a day or two after you create it.
00:11:57 If you want to take a look at how the math is working out or maybe you'd like to change the
math, you just simply
00:12:03 select the cell in the column here in the header. And then enable the formula bar and then you
can see that.
00:12:13 And you'll see that like B2, C2, D2, and E2. In this case, these just refer to the line here.
00:12:22 So we essentially have A, and they're going B, C, D, and E. So when we sum this up, it'll be the
first three quarters
00:12:32 of actuals and the last quarter for our forecast. So with this, I would say have fun and explore
00:12:39 the exercises that are provided, but feel free to do some self exploration as well. That's part of
the process of learning.
00:12:49 And really, I encourage you to have fun. Thank you.

28
www.sap.com/contactsap

© 2020 SAP SE or an SAP affiliate company. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or for any purpose without the express permission of SAP SE or an SAP affiliate company.

The information contained herein may be changed without prior notice. Some software products marketed by SAP SE and its distributors contain proprietary software components of other software vendors.
National product specifications may vary.

These materials are provided by SAP SE or an SAP affiliate company for informational purposes only, without representation or warranty of any kind, and SAP or its affiliated companies shall not be liable
for errors or omissions with respect to the materials. The only warranties for SAP or SAP affiliate company products and services are those that are set forth in the express warranty statements
accompanying such products and services, if any. Nothing herein should be construed as constituting an additional warranty.

In particular SAP SE or its affiliated companies have no obligation to pursue any course of business outlined in this document or any related presentation, or to develop or release any functionality
mentioned therein. This document, or any related presentation, and SAP SE’s or its affiliated companies’ strategy and possible future developments, products, and/or platform directions and functionality are
all subject to change and may be changed by SAP SE or its affiliated companies at any time for any reason without notice. The information in this document is not a commitment, promise, or legal obligation
to deliver any material, code, or functionality. All forward-looking statements are subject to various risks and uncertainties that could cause actual results to differ materially from expectations. Readers are
cautioned not to place undue reliance on these forward-looking statements, and they should not be relied upon in making purchasing decisions.

SAP and other SAP products and services mentioned herein as well as their respective logos are trademarks or registered trademarks of SAP SE (or an SAP affiliate company) in Germany and other
countries. All other product and service names mentioned are the trademarks of their respective companies. See www.sap.com/copyright for additional trademark information and notices.