1
Buy extended maintenance
Costs a surcharge and buys three years of breathing room. After that, the same decision is due again, only with less time to make it.
SAP Analytics Cloud · Planning and reporting
Your planning needs a new home. By the end of 2027.
On 31 December 2027, mainstream maintenance ends twice over: for Business Suite 7 and for the planning tools alongside it, BPC and BW 7.5. Anyone planning there today therefore loses not one tool but two foundations on the same day. The decision is usually made late, as part of the ERP project, and the business department rarely has a seat at the table. Then the new platform is in place, and planning is the last thing to move.
Coretica keeps functional design, data model and implementation in one hand. You explain your planning once. What ends up in the system afterwards was built by the same person you talked it through with.
The deadline
Two endings, one date
14 months.
until the end of mainstream maintenance on .
The end of maintenance is not a shutdown. It is the day from which there are no more fixes and no more legal updates. For a planning solution that carries closing figures, that amounts to the same thing.
Mainstream maintenance
Extended maintenance
-
30 Jun 2026
BPC 10.1 for Microsoft, already ended
-
31 Dec 2027
Business Suite 7, BPC 10.1 for NetWeaver, BPC 2021 on BW/4HANA 2021, BW 7.5
-
31 Dec 2030
End of extended maintenance, BPC 2021 on BW/4HANA 2023
For S/4HANA and BW/4HANA, by contrast, maintenance is committed until at least 2040. Moving now does not mean moving into the next stopgap.
BPC only follows this path together with the foundation beneath it: BPC 2021 on BW/4HANA 2021 is maintained until the end of 2027, on BW/4HANA 2023 until the end of 2030. As of mid-2025, none of the sources gave a date for later versions. The authoritative reference remains SAP's Product Availability Matrix for your own version.
Sources: Maintenance phases of Business Suite 7 according to cmt.de · BPC dates and SAP Analytics Cloud as the strategic planning solution according to the SAP Community · BW 7.5 according to IBsolution. Retrieved on 7 September 2026. BPC 2021 per BW/4HANA version according to Concentric Solutions and Swap Support. Retrieved on 14 October 2026.
Sources for the views: Compass, Monte Carlo simulation, ranges and number of runs according to FIS · Lesson “Simulating data with SAP Analytics Compass” in the SAP learning platform · Task types, dependencies and notifications in the planning calendar according to BIX Consulting · “Simulation (Compass)” and the calendar task types in SAP Help, version Q3 2026 · Import limit according to SAP Note 2882683. Retrieved on 7 September 2026.
The decision
Three ways out of the same deadline
2
Move planning to SAP Analytics Cloud
One model, one interface, the same data foundation as the new ERP. The move is the work. After that it is done, and the deadline is no longer one.
The path Coretica builds
3
Put a third-party tool alongside
It works, and for some companies it is the right choice. But it costs a second integration path, and the financial data then sits in two places.
In the system
Ten views into a planning solution we built
Ten views from the same solution. They are connected: the calendar assigns the task and triggers what runs without anyone stepping in, the form captures the figure, the posting logic carries the calculation forward, Just Ask queries it, Smart Insights breaks down the variance, Smart Predict suggests the next value, Compass runs the entire plan ten thousand times, and whatever comes out of it, the business department reads in its own view.
Planning cycle Budget 2027
from template “Quarterly forecast” · 14 tasks, 3 of them automatic
Volume planBusiness department · Input
completed
Cost ratesControlling · Input
completed
Headcount ramp-upHR · Input
in progress
Project budgetsProject management · Input
deadline running
ApprovalDivision management · Review task
open
Calculate forecastData Action · automatic
runs 17 Oct
Lock versionData locking · automatic
after approval
completed in progress deadline running runs automatically
The planning cycle is assigned in the system, not via a mailing list. Every task has an owner, a deadline and a reviewer, and tasks can depend on one another. Nothing is rebuilt for the next cycle: a calendar template produces the same process with the same tasks, deadlines and owners. Three steps run without anyone stepping in: the calendar starts the Data Action that calculates the forecast, it pulls the version and locks it after approval.
Budget 2027 (private)
StructureActual 2026Budget 2027ForecastΔ
Production division1,240
1,1801,205−2.8%
4210 Assembly720
690704
−2.2%
Line 118 · M. Berger, 4 Sep430
412420−2.3%
Line 119 · M. Berger, 4 Sep290
278284−2.1%
4220 Surface finishing520
490501
−3.7%
Line 204 · S. Lang, 5 Sep520
490501−3.7%
Planning is done in a private version. It is only published on approval, and until then nobody else sees the figures. The total at the top is calculated from the lines beneath it, right down to the single one a planner entered.
Data Action · Advanced Formulas
// Allocation of administrative overheads by person-days.
// Simplified excerpt, sample figures.
MEMBERSET [d/Version] = ("Budget2027")
FOREACH [d/CostCentre]
DATA([d/Measure] = "ALLOC_ADMIN") =
RESULTLOOKUP([d/Measure] = "OH_ADMIN",
[d/CostCentre] = "#")
* RESULTLOOKUP([d/Measure] = "PERSON_DAYS")
/ RESULTLOOKUP([d/Measure] = "PERSON_DAYS",
[d/CostCentre] = "#")
ENDFOR- Load actuals
- Calculate allocation
- Apply cost rates
- Copy and lock version
- Notify participants
Simplified excerpt. The script is one step; above it sits the chain that an owner triggers with a single click. The rule lives in the model, not in a cell, and it calculates the same way for every cost centre.
Workbook · Budget 2027
Cost centreBudget 2027Status
4210 Assembly690sent
4220 Surface finishing490sent
4230 Testing and trials1,940rejected
4240 Final inspection305sent
The value exceeds the approved limit and will not be written to the model. Please enter a justification or request approval.
Anyone at home in Excel stays in Excel. The cells are linked to the model, not to a file; permissions, locks and validation rules apply just as they do in the web interface.
Just Ask · Question to the model
Which cost centres are over budget in the third quarter, and why?
Answer
4230 Testing and trials+184
4310 Maintenance+96
4520 Logistics+41
Variance from budget in thousand euros. Every figure can be expanded down to the single line, including the planner’s initials and date.
What these tools deliver is a pointer, not a decision. They show where to look; the variance is assessed by whoever is responsible for the cost centre.
Smart Insights · Where the variance comes from
determined automatically
Testing and trials is €184k over budget in the third quarter. The system breaks down the variance without anyone having to search for it.
Test rigs, person-days
+121
Maintenance, materials
+38
External services
+25
Travel expenses
+7
Other, offsetting
−7
Total variance
+184
Two thirds of the variance sit in a single item. Getting this far used to take half a day of searching and ended in a spreadsheet nobody could recalculate. Every bar can be expanded down to the posting.
Smart Predict · Forecast suggestion Q4
time series model, 80 per cent range
MonthSuggestionRangeSetby
October10496 to 112
104M. Berger, 5 Sep
November10193 to 109
118S. Lang, 5 Sep
December112104 to 120
open—
Values in thousand euros. November is above the range: the planner knows about a special production run that past data cannot reveal.
The planner does not get an empty field but a calculated suggestion including a range, and corrects it where they know better. The model always holds the value a person has confirmed.
Compass · Monte Carlo scenario simulation
10,000 runs
Instead of a single figure, the assumptions are entered as a range. The system runs the plan ten thousand times and shows how the result is distributed.
AssumptionfromtoEffect on the result
Inflation on material costs3.0%
5.0%high
Sales volume−8%
+6%high
Personnel costs+2.0%
+4.5%medium
External services−3%
+9%low
90 out of 100 runs fall within this range. Operating result 2027 in €m.
The result is not a figure but a statement about probability. That is exactly the language a supervisory board expects when it asks about risk. In SAP Help, Compass is listed in the planning part of the solution and requires a planning licence; a pure BI authorisation is not enough. The Q1/2025 release is named by the source below, not by SAP Help itself.
Story · Production division
Budget 20271,180
Forecast1,205
Variance+25
Lines in division118
Budget Forecast in thousand euros
Every division gets its own view of the same model. No second data set is created that someone then has to reconcile.
Simulation · Version “What if”
What happens if inflation on material costs rises from 3 to 5 per cent?
Assumption
Inflation on material costs
3.0%5.0%
Collective pay increase, staff
+2.8%
Affected item
Material costs
3,1803,242
Personnel costs
6,665
Result
9,907
Total costs 2027, in thousand euros
previously 9,845
Effect on costs +62 0.6 per cent more, visible before approval
The assumption is changed on the left, the effect appears on the right, without anyone having to update a spreadsheet. Anyone who needs a range instead of a single assumption calculates the same question with Compass and gets a probability instead of a figure.
Our own illustrations in Coretica design, built to reflect how SAP Analytics Cloud works. Sample figures; no data from a client system and no screenshots of an SAP interface.
Day to day
What changes day to day
Six areas in which the work in controlling shifts. For each area: how it usually runs today, what SAP Analytics Cloud changes about it, which effort disappears and what needs to be settled beforehand.
Analysis and reporting
- Today without SAC
- The monthly report is built from exports that someone copies together and reconciles against each other.
- With SAC
- Reports read directly from the model that planning also happens in. Each area has its own view of the same figures.
- Where effort disappears
- In gathering data before every report and in asking which version is the valid one.
- Prerequisite
- Agreed master data and hierarchies that all areas refer to.
In the system: Business department view
Planning
- Today without SAC
- Planning files go out to the departments by email and come back in different versions.
- With SAC
- Data is entered in planning input forms with validation rules, initially in a private version. Anyone who prefers to work in Excel writes from there into the same model.
- Where effort disappears
- In collecting and consolidating files and in sending faulty ones back.
- Prerequisite
- An agreed planning structure: who plans what, at which level and with which assumptions.
In the system: Planning input form · Excel as an input form · Posting logic
Workflow and coordination
- Today without SAC
- Deadlines live in emails. You find out who has not delivered yet by chasing people.
- With SAC
- The calendar assigns tasks with an owner, a deadline and a reviewer. Approval is a review task, a data lock then freezes the version, and the next cycle is created from a calendar template.
- Where effort disappears
- In reminder emails, ring-rounds and the separate list in which someone tracks the status.
- Prerequisite
- A process with roles and deadlines that the departments have agreed on beforehand. The system maps it; it is not invented there.
In the system: Calendar
Simulation
- Today without SAC
- Three scenarios by hand, with no statement about how likely they are.
- With SAC
- Compass recalculates the plan many times using the Monte Carlo method. The result is a range with probabilities instead of a single point value, and the driver model shows which assumptions carry the result.
- Where effort disappears
- In the copies of the plan that are maintained for each scenario.
- Prerequisite
- A model in which drivers and result are linked by formulas, and a planning licence.
More on this: Monte Carlo simulation · In the system: Compass · Drivers and simulation
Artificial intelligence
- Today without SAC
- Explaining a variance means filtering, searching and asking the department.
- With SAC
- Just Ask answers questions asked in full sentences, Smart Insights shows the factors behind a variance, Smart Predict proposes a forecast with a range. Joule, too, currently comes into play in SAC via Just Ask.
- Where effort disappears
- In searching for the cause and in facing an empty forecast field.
- Prerequisite
- The functions depend on data centre, activation and licence. Whether your tenant can use them is something we clarify in the project, before anyone plans with them.
In the system: Just Ask · Smart Insights · Smart Predict · More on this: AI in planning
Integration and security
- Today without SAC
- Figures are exported, distributed by email and then sit in many places, each with its own status.
- With SAC
- Actuals from S/4HANA, BW/4HANA and non-SAP sources flow into the same model, via SAP Datasphere for large volumes. Authorisations go down to the individual figure, and data locks protect approved versions.
- Where effort disappears
- In exports, file shares and the question of who may see which figure.
- Prerequisite
- Clarified roles and access rights, and an architecture that fits the data volume.
More on this: The layer underneath
Simulation
Monte Carlo simulation: how robust a plan is
A plan with a single number conceals how certain it is. A Monte Carlo simulation draws the assumptions at random many times, calculates the whole plan each time and collects the results into a distribution. SAP Analytics Cloud comes with Compass for this.
-
A range for each driver.
Instead of one number, a minimum and a maximum value. Values are drawn at random in between.
- Sales volume
- Inflation on material costs
- Personnel costs
-
The whole plan calculated ten thousand times.
Each line stands for ten runs. Many results add up to a distribution.
-
The target gets a probability.
In 16% of the runs the target of €7.5m is missed. 90% lie between €7.2m and €8.7m.
What goes in
A minimum and a maximum value for each driver. In between, Compass draws the values from a normal or a uniform distribution, as selected. Anyone who wants to limit individual drivers sets them up as Restricted Drivers.
How it is calculated
You can choose 1,000, 10,000 or 100,000 runs. The basis is the formula in the model that links the drivers to the target figure. More runs take longer and give a smoother picture.
What comes out
A distribution of the result. It shows the range in which most of the runs ended and how often the target was missed. Together with the driver model, it shows which assumption moves the result most and where more precise planning pays off.
What it requires
SAP Analytics Cloud for Planning, licensed at the Professional or Standard tier. In addition, a driver-based model and ranges that someone can justify, for example from prior years or the order backlog.
An example with sample figures is shown in the Compass tab. How to read the result and where simulation does not help is explained in the article Monte Carlo in planning.
Source: Lesson “Simulating Data with SAP Analytics Cloud Compass” in the course “Leveraging Advanced Features in SAP Analytics Cloud for Planning”, SAP Learning. Read on 14 October 2026.
Building blocks
Five solutions, individually or together
Each of these solutions can be started on its own. All five work on the same data model, so the second one does not start from scratch.
Financial planning and forecasting
Cost, profit and budget planning with input forms, validation rules and separate versions for plan, forecast and actuals. The forecast continues on the same data instead of being built in a second file.
Capital expenditure planning
Investment projects from request through approval to drawdown, with budgets and the follow-up costs attached to them. This is the building block the group solution started with.
Workforce and resource planning
Who is committed to which project and when is shown in one view and linked to the personnel costs in financial planning. Bottlenecks and free capacity become visible before they turn into costs.
Management reporting
Reports for management on the same figures that planning uses. Variances can be traced from the total down to the individual posting.
Simulation and scenarios
What-if calculations via driver trees. Where assumptions are uncertain, Compass runs a Monte Carlo simulation. The result is a range with probabilities instead of a single value.
Costs and benefits
What SAC costs and where it pays off
There are three items on the cost side. Anyone who knows them before the decision will not be surprised during the project.
Licence
The licence comes from SAP. Which tier is needed depends on the functions: planning and Compass require a planning licence. SAP publishes the prices itself, and they change. That is why there is no figure here.
Implementation
Functional concept, data model, implementation and training. The effort depends on the number of planning areas, data sources and people involved. We state scope and price after the Efficiency Check.
Operation
Master data and the model need maintenance, and new requirements come along. Your business departments take this on themselves after training, or we support it to an agreed extent.
On the other side: the hours in the process
The benefit arises where data is collected, reconciled and corrected today. The home page lists five levers for this:
- One data basis instead of reconciliation
- Validated input instead of after-the-fact correction
- Reminders and deadlines in the system instead of by email
- Ranges from simulation instead of three manual scenarios
- Reports at a click instead of copying together
There is no universally valid benefit figure for SAC. We calculate it for your case.
The layer underneath
When planning outgrows its model
Planning grows with the company. At some point it no longer fits into an import model in SAC, and then the question arises of where the data will live in future. This is not a side issue for IT: it decides whether the planning will still run in five years.
Where the data lives
SAP Analytics Cloud
Planning, reporting, simulation: the interface for the business department
SAP Datasphere
the persistence layer, now part of SAP Business Data Cloud
S/4HANA
actuals, master data
BW/4HANA
prepared financial data
Non-SAP sources
upstream systems and files
An ordinary live connection to a Datasphere model does not support planning. There is a separate mechanism for that, which SAP made generally available in 2025. Anyone who confuses the two spends a year planning on the wrong architecture.
Coretica has carried out exactly this move itself for a group-wide planning application whose data volume had exceeded the import model.
Artificial intelligence
What AI really does in planning
In planning, it takes over the three tasks that cost planners the most time: filling the empty field, tracking down a variance and requesting an analysis that does not exist yet. In the processes alongside, SAP has long gone a step further and lets agents take over entire transactions.
Just Ask Smart Insights Smart Predict Joule, being expanded
Joule · assistant and agent
checked on 7 September 2026
Assistant · informs
You ask, it answers: a figure, an explanation, the way to the right view. This is how Joule works in Analytics Cloud today, via Just Ask.
Agent · executes
It takes over a transaction: checks, decides within the set framework and posts. Reports speak of more than 40 such agents with around 2,400 skills, spread across finance, supply chain, procurement and HR.
- Creating and justifying journal entries
- Cutting invoice dispute resolution from days to hours
- Proposing provisions
- Checking production release against material and capacity
This is the difference that gets lost in presentations: an assistant informs, an agent acts. Joule Studio can be used to build your own agents for your own business rules. Today’s agents, however, sit in the processes and not in planning. For a planning landscape this means: the benefit comes first from asking questions of the data; process automation follows.
What Joule requires
- a shared identity service across all systems involved: Joule and every connected SAP product on the same Cloud Identity Services tenant, with a uniform user ID. In landscapes that have grown over time, that alone is a project of its own.
- a global BTP enterprise account with the Joule entitlement in the Foundation plan, plus SAP Build Work Zone in the same account
- SAP AI Core in the extended plan, not in the standard plan
- the licence for the respective application, plus AI Units for the paid features
So the question in a project is rarely whether Joule can do something. It is what your system landscape brings to the table for it, and in which order the gaps should be closed.
How the parts fit together
checked on 7 September 2026
Joule
SAP’s assistant across applications, currently works in SAC via Just Ask
Just Ask
ask instead of search
Smart Insights
cause instead of guesswork
Smart Predict
proposal instead of an empty field
Your planning model
Actuals, plan and forecast, hierarchies, authorisations, locks
All three access the same model. That is why the quality of the model determines what an assistant can answer at all, and not the number of AI features in the brochure.
In the platform today
- Just Ask
- Questions in full sentences, answers as a figure or chart, expandable down to the row.
- Smart Insights
- Shows the influencing factors for a variance.
- Smart Predict
- Calculates a forecast proposal with a range from past data.
Which of these functions a specific tenant can use depends on data centre, activation and licence. This is checked in the project and not assumed. And one thing remains: a proposal is a proposal. The value in the model is the one a person has confirmed. The statements on Joule were checked against the SAP documentation on 7 September 2026: assistant and agent, Joule Studio, the shared identity service and the Foundation plan are stated there. The number of agents and their skills comes from a public summary by an SAP partner, not from a commitment by SAP.
After the project
Analysing without building every time
Analysis at the push of a button
Whatever is in the model can be analysed without anyone first building a data extract.
Only what belongs to the question
Filters on every chart, from the total down to the individual posting, in the same view.
A separate view for each business department
Each area has its own story on the same model. No second data set that would need reconciling.
History remains available for analysis
Everything collected stays in the model. You do not have to analyse it, but the question of the day after tomorrow can reach today’s data.
Same data, different questions
Sales planning, workforce ramp-up, capital expenditure calculation, project control, simulations. What has been modelled cleanly once carries the next use case, without buying a new system.
Each group sees its own form
Input forms, KPIs and authorisations tailored to each role. The plant manager does not get the same interface as group controlling, and both work on the same figures.
Origin
Built where it is hardest
- Several key metrics in one model, not in several side by side.
- Actuals from SAP and non-SAP sources, in the same structure.
- Planning cycles with their own deadlines and locks that do not block each other.
- Authorisations down to the individual figure.
- Rollout across a complete vehicle project, with expansion to further areas under way; subsequent operation by the business departments themselves.
The solution became the standard there for the planning processes concerned, and further projects are actively driving its use and its expansion.
How such an assignment runs from the initial situation to the result is described on a separate page: the reference project in detail.
Questions
Questions and answers
Which consultancy supports SAP Analytics Cloud projects in Munich?
Coretica is based in Munich and advises on planning and reporting with SAP Analytics Cloud, from the functional concept through the data model to implementation. Workshops and alignment meetings take place on site at your premises; most of the build is done remotely. We take on projects throughout Germany.
What does implementing SAP Analytics Cloud cost?
The costs consist of the SAP licence, the implementation and ongoing operation. SAP publishes licence prices itself, they change, and that is why they are not given here as a figure. The effort of implementation depends on the number of planning areas, data sources and people involved; we state a reliable order of magnitude after the Efficiency Check. More on this in the section Costs and benefits.
Is SAC worthwhile for mid-sized companies?
This is decided less by the size of the company than by the effort that currently sits between the steps of planning and reporting. A mid-sized company can start with a single building block, such as financial planning, and expand later. Whether it pays off is shown by the benefit calculator with your own assumptions.
What happens to BPC and BW planning after the end of mainstream maintenance in 2027?
It keeps running, but no longer receives corrections and legal updates from standard maintenance. For BPC 10.1 and BW 7.5 there is extended maintenance until 2030 for a surcharge, and BPC 2021 on BW/4HANA 2023 is maintained until the end of 2030. After that, the same decision is due: move the planning, for example to SAP Analytics Cloud, or put another tool alongside it. Details in the article End of maintenance 2027.
What does a Monte Carlo simulation bring to planning?
Instead of a single plan value, you get a distribution that shows in which range the result lies and with what probability. To do this, each uncertain assumption is given a range, and the plan is recalculated thousands of times. In SAC this is done by Compass; how is described in the section Monte Carlo simulation.
Can existing systems and Excel still be used?
Yes. Actuals come from the systems that already exist, from S/4HANA as well as from upstream systems and files. Anyone at home in Excel keeps entering data there, but then writes into the planning model instead of a separate file. What gets replaced are the copies and reconciliations in between, not the tools that work well.
Do you offer SAP Analytics Cloud training?
Yes, training is part of every project. Training takes place on your own model and not on a textbook example: planners in data entry, controlling in reports and analyses, those responsible in operating the model. The aim is that your business departments manage the solution themselves afterwards.
How long does it take to implement SAC planning?
That depends on the scope. A single planning building block with few data sources is a different task from group-wide planning with many people involved. It starts with a current-state assessment, and after that we give a duration that is based on your processes and not guessed.
SAP Analytics Cloud or BPC: what is the difference?
BPC is a tool for planning and consolidation that, depending on the version, runs on SAP BW, BW/4HANA or S/4HANA. SAP Analytics Cloud is a cloud service that combines planning, reporting and analysis in one interface, and SAP positions it as its strategic planning solution. For consolidation, SAP names a different target product.
How is capital expenditure (CapEx) planning mapped in SAC?
As a planning area of its own: investment projects are requested, reviewed, approved and drawn down, together with budgets and follow-up costs such as depreciation. Approvals run as tasks in the calendar, and the figures flow into the financial planning of the same model. The group solution in our reference project started with exactly this building block.
Is SAP Analytics Cloud itself affected by the end of maintenance?
Technically no, in terms of timing yes. SAC is a separate solution with its own lifecycle. Its data source, however, is affected by the end of maintenance, and that is why the decision about both is made together.
Isn’t this scaremongering?
No. An end of maintenance is not a shutdown, and extended maintenance is a viable path. It is just a path with a surcharge and with the same conversation three years later.
Why not simply stay in Excel?
Because there the question of which version is valid is never answered, the rules sit in cells instead of in the model, and the knowledge about them depends on one person. As long as that goes well, it goes well.
Who works on our project?
Architecture, functional steering and the critical building blocks are handled by the founder. For larger projects, the team is extended on a project basis with vetted specialists, after agreement and by name; steering stays in one pair of hands.
That is the difference from a large consultancy, and it is intentional: one contact who builds the solution personally, instead of a tiered set-up in which the experience is in the proposal and someone else sits in the project.
When can you start with us?
That depends on the scope. No two planning projects are alike, and scope, effort and scheduling are assessed before we commit.
A current-state assessment with a target picture can usually be scheduled at short notice; an implementation with rollout is scheduled together. Write to us with one sentence about what it is about, and you will get a reliable answer instead of a ballpark figure.
Tell us what your planning runs on today.
One conversation is enough to see whether the move is a matter of weeks or of months for you.
Further reading
Articles on planning in SAP Analytics Cloud
-
Replacing BPC: how planning moves to SAP Analytics Cloud
What cannot simply be transferred, the steps in which we migrate and what migrations usually hinge on.
-
Capital expenditure planning in SAP Analytics Cloud: from request to depreciation
The path of a project through request, approval, drawdown and profit planning in one model.
-
End of maintenance 2027: what happens to BPC and BW planning now
What the date means, which three paths exist and what should be clarified before the decision.
-
From scattered Excel files to one figure: the path to a single point of truth
Why two departments have two values, and six steps to a shared data basis.
-
Monte Carlo in planning: ranges instead of three scenarios
How a simulation turns uncertain drivers into a distribution and what it requires.