Power BI and Microsoft Fabric

Reports to which the department believing

We build Power BI environments for SMEs – with a coordinated data model behind them, clear responsibilities and a licensing model that fits your number of users instead of the price list.

A model instead of twelve truths
Licences in advance
The question on which the costs depend
Per user or central capacity?
Per user
Everyone who watches or creates reports needs their own license. Easy to calculate, but increases linearly with the number of readers.
Central capacity
They buy computing power. Viewers then do not need their own paid license – this is clearly worthwhile from a certain number of readers.
Where exactly the turning point lies depends on the number of pure observers – and this is almost always underestimated.

What companies come to us with

In the rarest cases, it is about a missing diagram. This is one of four situations.

Two reports, two figuresSales and controlling call different sales, and both can justify it. This is not a tool problem, but a lack of a common data model.
Reporting lives in ExcelSomeone assembles the same evaluation by hand every month. It works – until the person fails or the file becomes too large.
The license issue is unresolvedShould each reader get their own license or is it worth a central capacity? The answer determines a significant part of the ongoing costs.
“Do we need Fabric now?”Microsoft is visibly expanding Fabric, and the question comes reliably from the management. It can be answered – but not in general.

Power BI or Fabric – when the jump is worth it

Fabric is not a new Power BI, but the platform below it: storage, data preparation and evaluation in an environment. For many medium-sized tasks, this is more than is needed – just right for some.

Power BI is enough
The more frequent case in the middle class
  • The data comes from a few, easily accessible sources – ERP, CRM, a database.
  • It is sufficient if data is updated at night or several times a day.
  • The data sets move in a size that a classical model can handle.
  • There is no team that could or wanted to maintain a platform permanently.
Fabric is worth it
If several points apply
  • Many heterogeneous sources should converge in one place, even unstructured or semi-structured ones.
  • There are already data paths drifting apart, and no one has the overview.
  • Data science or machine learning should work on the same data as reporting.
  • The number of users is so high that a capacity license would be cheaper anyway.

Our experience: The change rarely fails because of the technology and often because nobody takes care of the new platform. Anyone who introduces Fabric without clarifying responsibilities exchanges a clear problem for a larger one. What happened last is in the article Power BI and Fabric become agents.

Licenses: per user or capacity

The license logic is manageable once you understand the one distinction – and expensive if you meet it too late.

Pro (per user)
approximately € 12 per user and month
The standard for creating and sharing. Anyone who watches reports also needs a license. In larger Microsoft 365 packages, Pro is already included – this is often overlooked and paid twice.
Premium per user
approximately € 21 per user and month
For individuals who need advanced functions – more frequent updates, larger models. Useful for a small data team, not for breadth.
Fabric capacity
Billing via Azure
You buy computing power instead of user licenses. Viewers then no longer need their own paid license. Bookable depending on use or as a reservation – and pauseable, which helps with plannable loads.

The rule of thumb: As long as only a manageable number of people view reports at all, user licenses are cheaper and easier. As soon as the reporting is to go broad – plant management, field service, shift manager – the bill tips towards capacity, because pure viewers do not cost anything extra there. Exactly this number of viewers is almost always set too low in estimates.

Information as guidance, as of August 2026, Microsoft list prices without discounts. Licensing models and prices change; For a reliable calculation, we check the valid status at the time of decision.

Why Two Reports Show Different Numbers

The most common reason is not a calculation error, but that each report has its own definition. one counts orders by date of receipt, the other by invoice; One takes out cancellations, the other does not. Both are right on their own, and that is why the discussion is so tough.

The solution is a common data model, in which key figures are once defined and maintained centrally. Reports use it instead of bringing their own logic. This does not take anything away from the specialist department – it can continue to evaluate freely, but on a basis that no longer needs to be disputed.

This includes an order that does not set itself up: who is allowed to publish and who only read which reports are considered released, what happens to the many drafts that emerge over time. Without these rules, Power BI creates the same wild growth you wanted to get rid of with Excel – only more colorful.

We set up this structure and ensure that the database behind it is right. How the routes look like is on our side Business Intelligence and ETL; what old routes fail in the contribution to ETL lines no longer touched.

Frequent questions about Power BI and Fabric

In most medium-sized cases, not at first. Fabric is worthwhile if many heterogeneous sources are to converge, if data science works on the same data or if the number of users suggests a capacity license anyway. If these points are not true, a clean power BI environment with a good data model is the cheaper and lower-maintenance solution. We tell you openly if the jump is not worth it.
There are two ways: licenses per user or a central capacity. The standard per user is around 12 euros monthly, the extended variant around 21 euros (list prices, as of August 2026). A capacity is billed via Azure; Pure viewers do not need their own paid license. Which way is cheaper depends mainly on the number of people who only watch reports – which is usually underestimated. We calculate both variants with your real numbers.
Partly. Some larger Microsoft 365 packages already include the Pro license. This is regularly overlooked and then paid a second time. A look into your existing licensing therefore belongs at the beginning of every calculation.
For individual evaluations, yes, and we do not advise anyone to abolish working tables. It becomes critical if a monthly evaluation is assembled by hand, hangs on one person or if several versions of the same file circulate. Then Excel is not the problem, but that it is the only place where logic exists.
The location of the Power BI environment is set up; Regions in the EU are available. Where data is not allowed to leave the house, we work with a connection to local sources instead of a complete relocation. This is a question we clarify before construction, not afterwards.
Yes. At the beginning, there is an inventory of which reports are actually used, where key figures are defined twice and differently, how rights were granted and what data links lie behind them. Experience shows that a significant part of the existing reports are orphaned – Cleaning up often brings more than any new evaluation.
ALGEBRA Engineering & Consulting GmbH
Power BI and Fabric for SMEs – Stuttgart location, activity nationwide.
Ulmerstraße 300
70327 Stuttgart
What a First Talk Clarifies
  • How Many People Will Really Watch Reports
  • Which key figures are controversial – and why
  • Whether your existing licensing already covers something
  • Whether Power BI is enough or Fabric is worth it

Settlement of talks

A number on which everyone agrees
We look at your reports, your number of users and your license situation and tell you what’s worth it – even if the answer is that you don’t need anything new.

Contact now