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.
Licences in advance
What companies come to us with
In the rarest cases, it is about a missing diagram. This is one of four situations.
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.
- 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.
- 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.
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
70327 Stuttgart
- 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