Getting a Power BI dashboard built: how it works

Getting a Power BI dashboard built usually costs between €1,500 and €10,000, and the first working version is ready within one to two weeks. Where you land in that range depends on your data sources and how complex your data model needs to be, but mostly on something most quotes skip entirely: whether anyone has actually decided which decision that dashboard needs to improve.

Skip that question, and the pattern is predictable: a sleek dashboard, a great first impression at the kickoff meeting, and a few weeks later nobody logs in to look anymore. Sound familiar? Not because the dashboard is technically bad, but because nobody defined upfront what was supposed to change because of those numbers.

What exactly is a Power BI dashboard?

A Power BI dashboard is an interactive screen that automatically pulls data from your source systems, think accounting software, CRM, webshop or Excel, and turns it into charts, tables and kpi tiles. The difference with an Excel report is fundamental: that is a snapshot made by hand, and by the time it is finished the numbers are outdated. A dashboard refreshes itself.

So what is a dashboard, without the brand name? The same idea minus Microsoft: one screen that keeps itself up to date and shows how things stand. It is also not a report. A report explains what happened and why. A dashboard shows at a glance how things stand right now. You need both, they just answer a different question.

The word you will meet in quotes for the craft around it is dashboarding. What is dashboarding exactly? It is the whole of opening up sources, building a data model, defining kpis and keeping all of that running. The dashboard is the visible end product, dashboarding is the work underneath it. That distinction is not wordplay, because you pay for the second and judge the first. Dashboarding for a business is rarely about the charts, and almost always about the agreements and the model behind them.

What does it cost to get a Power BI dashboard built?

Having a dashboard developed is not a fixed price. It ranges from around €1,500 for a simple dashboard with one data source to €10,000 or more for a custom-built dashboard with multiple systems and permissions per department. Three things determine where you land in that range:

• Number and type of data sources. Connecting one Excel export is an hour’s work. A connection to AFAS, Exact or a CRM system needs a stable, repeatable link.
• Complexity of the data model. A dashboard that only shows monthly revenue is simple. Tying revenue, margin and stock together needs a well thought-out data model: work you never see on screen, but that decides whether the dashboard is correct.
• Number of users and roles. One report for the board is a different job than a dashboard where twelve locations each see their own numbers, with the right access per user.

The comparison that counts is not only what the build costs, but what it costs afterwards: billing every change separately adds up, and the Power BI licensing costs per user land on top of that.

What goes on a KPI dashboard?

A KPI dashboard is not a different product from the dashboard above, it is how you fill it in: the five to eight figures a decision hangs on, each with its target beside it. Without that target you are reading a position rather than a signal.

Here is what such a dashboard example looks like in practice, for a wholesaler with one location:

• Four tiles at the top. Revenue against budget, gross margin as a percentage, outstanding receivables in days, and the share of orders shipped complete. Each with the target beside it and the gap underneath.
• Two charts below that. Revenue per month against last year, and margin per product group. This is the layer that answers the question of how that happened.
• One table at the bottom. The ten largest customers with their revenue and margin, drilling through to the individual orders.

A first dashboard does not need to be more than that. The temptation is to put everything measurable on it, and that is exactly how you end up with a screen nobody takes anything from. Which KPIs it should be in your case depends on the decision that has to get better, which is the next section.

How does a Power BI implementation run?

In four phases, and each phase hands you something concrete. That is the question to ask in a quote: not how long it takes, but what is on the table at the end of each phase.

• Analysis, a few days. Which decision has to get better, which figures determine it, and where they sit today. This produces a list of kpis with a definition and a source for each.
• Foundation, weeks one and two. Connecting sources and building the data model. This produces the first working version, with your own figures in it.
• Expansion, weeks two to six. Extra sources, role-based permissions, and refinement based on actual use. This produces the dashboard as it will keep running.
• Handover. Your own team learns to adjust the model and build reports themselves. This produces the state where you no longer need us for every change.

So that first version arrives quickly, and not because speed is the goal. A team can only say something useful about a screen it can touch. Our approach is built on that order.

Which systems can you connect?

Power BI connects to almost any system that can export data: accounting software like Exact and AFAS, CRM systems, webshops and Excel. Standard connectors exist for most Dutch accounting and ERP packages; older or specific systems get connected through a custom link. What such a link looks like from the inside is set out step by step in our article on connecting AFAS to Power BI. If you run Exact Online, the three possible routes are in our article on connecting Exact Online to Power BI.

If you have several systems that each tell part of the story, stock in one package and sales in another, the data model decides whether Power BI ties those sources together correctly or whether you end up with two dashboards that contradict each other.

Once connected, the dashboard refreshes itself. With a Power BI Pro licence that can happen automatically up to eight times a day, without anyone needing to run an export. The official documentation on refreshing data in Power BI goes deeper into the technical side.

Start with the decision, not the dashboard

The first question most projects ask is about form: which colours, which charts, how many pages. We flip that order and ask about the outcome: which decision needs to turn out differently after this dashboard than it does now?

That answer differs by role, by organization, and sometimes by quarter. Which is exactly why you cannot skip it: a generic dashboard answers a generic question, and almost nobody is asking that one.

That is not just a house rule of ours. In its own adoption guidance for Power BI, Microsoft writes that BI solutions only deliver value once they align with the organization’s goals, and describes the organizations stuck at the level where the conversation is about features and tools rather than goals.

Only once that question is sharp do kpis, definitions and data logic follow, and only after that, Power BI. A dashboard that only summarizes yesterday just confirms what you already knew. If the steering is financial, the kpis that go with it are worked out in our article on a financial dashboard.

Build it yourself, have it built, or build it together?

Power BI Desktop is free, and making dashboards yourself works fine with some time. How you build a dashboard that still holds up six months later is a different question entirely. Where it usually goes wrong is not dragging a chart onto a canvas (anyone learns that within a day, and immediately afterwards everyone also thinks they are done), but the data model underneath: which tables relate how, which calculations still hold under a new filter, and why a report shows one person a different number than another.

That is the difference between building it yourself and having it built. Most clients pick the middle road: we build the foundation and the first version, your team builds alongside and takes it over. By the end your own team adds reports independently, and we stay available for the questions that genuinely matter.

Where it usually goes wrong at handover

Good Power BI dashboard design is not the same as a dashboard you can trust: a screen can look immaculate and still show the wrong figures. Four things go wrong most often, and all four can be checked at handover:

• The figures differ from the source. What is on screen should match your accounting. Nothing undermines trust faster than a dashboard that is “roughly right”.
• The data model is a pile of loose formulas. Tables that do not relate logically and calculations that each work out slightly differently. This is the work nobody sees, but it decides whether a new kpi can be added later in hours or in weeks.
• Nobody owns it. Who may make changes, and where the dashboard lives: in your own tenant under your own licences, or tucked away in an external party’s environment.
• It does not refresh by itself. The moment somebody still has to load an export, it is a report in different clothes.

---

Will this dashboard become something you glance at monthly, or an instrument you steer with weekly? Want to know what a Power BI dashboard would cost and return for your organization? Get in touch, we would rather work through a concrete example than talk in generalities.

Frequently asked questions