Business intelligence tools come in two broad models, and most of the confusion in buying one comes from not knowing which model you are choosing.
In the self-service model, you license a tool, such as Power BI, Tableau, Qlik or Looker, and your own people build and maintain the reports. In the managed model, a partner builds the reports, hosts the platform and keeps it running, and your people use the dashboards. Neither is better in general. Each fits a different kind of business.
What self-service does well
Self-service is strongest when you have people who want to explore. Analysts can follow a question wherever it leads, build a new view in an afternoon and share it with a colleague. If your business asks many new questions every week and you have two or three people who enjoy answering them, a self-service tool gives them room to work.
It also keeps knowledge in house. The people who understand the business are the ones building the reports.
Where self-service struggles
The cost of self-service is that you now run a BI function whether you meant to or not. Someone has to build the data model, maintain it as source systems change, manage who sees what, and keep the numbers consistent between reports. Without that discipline, a familiar pattern appears: several people build their own versions of revenue, the numbers disagree, and meetings start with an argument about whose figure is right.
Self-service also depends on a few key people. When the one person who built the model leaves or goes on leave, reports quietly stop being maintained.
What managed BI does well
Managed BI is strongest when most of your people need to read numbers, not build them. Consistency is built in, because the definitions are written once by one team. Access control is set up deliberately instead of growing report by report. The platform is somebody else's job to keep running, and reports get fixed when sources change.
It also suits situations with many readers and strict visibility rules: several plants, several companies, or clients who should each see only their own data.
Where managed BI struggles
The trade is flexibility. In a managed model, a brand new ad hoc question often goes through the partner instead of being answered by your analyst that afternoon. You also depend on the partner, which makes it important to ask hard questions before you start: who owns the definitions and the code, what documentation do you receive, and what are the terms if you want to leave.
If your business wants everyone to freely build their own analysis, managed BI is not the right model for them.
The hybrid most companies end up with
In practice many businesses land in the middle. A small number of analysts use a self-service tool to explore and prototype, while a managed platform serves the rest of the company with consistent, governed dashboards. The analysts find what is worth tracking, and the managed layer makes it reliable and visible to everyone.
That is a sound setup, and there is nothing wrong with using both.
How to protect yourself with a managed partner
If you choose the managed model, put the protections in writing at the start. You should own your definitions and any code built for you, or at least receive them in a usable form. You should receive documentation that lets another team understand what was built. Access to your source systems should be read-only and should be removed when the engagement ends. The terms for leaving should be clear, including how long you can keep using the platform while you transition.
A good partner will agree to all of this without fuss. One who resists is telling you something.
Questions that decide it
How many people will build reports, and how much of their week will they spend on it? How many people only need to read them? How costly is it when two reports show different numbers? Who will maintain everything in three years? How strict are your visibility rules? Honest answers to these usually point clearly in one direction.
We build in Power BI and Tableau for clients and we run Horizon, a managed platform, so we have no reason to push you toward either. We would rather you pick the model that fits. If you are still weighing tools, our guide to Power BI alternatives for mid-market companies goes through what to compare.
If you are deciding between the two and want an honest read on which suits your business, reach out.
Have a question about this topic? We work with businesses worldwide.
Get in Touch