# Management dashboards: why the numbers don't match > Your dashboard disagrees with finance because each department defines the metric differently and nobody built a single source of data. How to fix the order. - Author: Marlon Trettin - Published: 2026-07-16 · Updated: 2026-08-29 - Language: en - Canonical: https://yowpi.com/en/blog/management-dashboards-why-numbers-dont-match --- Your management dashboard does not match finance because the numbers come from different sources, calculated by rules nobody agreed on. The problem is not the BI tool or the chart. It is that the operation never had a single source of data. Before replacing the dashboard, define the metric in one place. ## TL;DR - Trust fell while the pressure rose. In [Salesforce research](https://www.salesforce.com/news/stories/trust-in-business-data-leaders-survey/) with 552 business leaders, confidence in data accuracy dropped 27% since 2023, and 76% say they feel more pressure to back decisions with data. - Fewer than half of leaders say their data strategy is aligned with business priorities, a 14-point fall since 2023. - The fix is architectural, not visual. Swapping the BI tool without a single source makes the wrong number prettier, and nothing else. - The agreement on what each number means costs meeting time, not budget — and it is the step almost everyone skips. ## Why doesn't the dashboard match finance? Because the dashboard's number and finance's number come from different places, under different rules. Sales counts a signed contract. Finance counts an issued invoice. The dashboard pulls a CRM extract that runs two days behind. Three true numbers, calculated three ways, that will never reconcile with each other. Leadership does not see that mechanism. It sees two screens that disagree and concludes, reasonably, that neither can be trusted. From that point the dashboard loses its function. It stops being the basis for a decision and becomes one more number to check in the meeting. The irritation is not a missing tool. It is a missing agreement on what each number means and where it comes from. Until that agreement exists, every department brings its own version and the meeting spends its good time arguing about which spreadsheet is right. ## Is the problem the BI tool or the data underneath? Almost always the data. Power BI, Looker and Metabase draw what they receive. If the source arrives inconsistent, the tool renders the inconsistency faster and with better styling. Southard Jones, chief product officer at Tableau, describes the real obstacle as fragmented enterprise data and complex analytics tools ([Salesforce](https://www.salesforce.com/news/stories/trust-in-business-data-leaders-survey/)). Changing the BI tool does not touch that. The data confirms the order of the problem. Fewer than half of leaders say their data strategy is aligned with business priorities, a fall of 14 points since 2023 ([Salesforce](https://www.salesforce.com/news/stories/trust-in-business-data-leaders-survey/)). The same logic holds for AI: without a base defined by people, a model just automates the same divergence faster. The obstacle appears before the screen, at the origin of the data. That does not absolve the tool of everything. A badly designed dashboard gets in the way even with good data. But the order of causation is clear: an inconsistent base defeats any tool, and no tool repairs an inconsistent base. ## What does a single source of data mean in practice? It means one place where each indicator is defined once, and every dashboard reads from there. It does not have to be an expensive data warehouse. In a smaller company it is usually a well-modeled operational database, fed by the systems that already exist, with each metric's definition written down and agreed between departments. The core of it is the definition, not the server. "Active customer" comes to have one meaning, not four. "Revenue for the month" comes to have a rule, written down, that sales and finance both sign off on. The single source holds the data; the agreed definition is what stops the number from diverging. When sales, finance and leadership read the same number, the conversation goes back to being about the decision. The argument over whose spreadsheet is correct disappears, and it should never have occupied a leadership meeting in the first place. ## What if the data is right and leadership still doesn't decide? Then the problem has moved from the base to the presentation. A dashboard that shows everything prioritizes nothing. A report can carry mountains of operational data without touching what actually matters to leadership, and that is exactly where it stops helping anyone decide. Leadership decides on three to five numbers, not thirty. Put those few at the top, with the comparison against target, and leave the rest in the detail, reachable by drill-down for whoever wants to go deeper. Good presentation hides the complexity without deleting the data. Measure how well the dashboard serves the decision, not admiration. If the meeting opens the dashboard and the first question is always "is this number right?", the problem is still the base. If the question is "what do we do about this?", the presentation did its job. ## How do you get out of a dashboard nobody trusts? Start with the numbers leadership actually uses to decide — three to five, not thirty. Write down the definition of each one and close the agreement between departments before touching any tool. That is meeting-and-minutes work, not technology, and it is the step almost everyone skips. Then map where each piece of data comes from today and reconcile the divergences. Here the real work shows up: discovering that sales and finance count the same sale at different moments is tedious, and it is exactly what has to be resolved. Only then build the single source that feeds the dashboard, and adjust the visualisation last. The order matters. Start with the screen and you rebuild everything when the numbers fail to reconcile in the first meeting. Start with the definition and you build once. ## A practical example The scenario below is hypothetical and does not describe a real client. A clinic group with 150 employees asks its innovation manager for a revenue-by-location dashboard. She builds it in Power BI, connects the extract from the scheduling system and presents it in the leadership meeting. The number does not match finance. The director asks why, and the meeting stalls on verification instead of deciding. Investigating afterwards, the cause surfaces: the dashboard counts scheduled appointments, finance counts invoiced appointments, and the scheduling export closes two days behind. Power BI was right. The definition of revenue was what had never been agreed. The way out was not to change the BI tool. It was to write down what counts as revenue, align that between scheduling and finance, and make the dashboard read from a single base. There is no glamour in it, and it is what gives the number its credibility back. Our cases follow the same logic. At [Reatop](/en/cases/reatop), hospital waste control moved from spreadsheets and paper notes, which delayed consolidation, to a real-time management dashboard that generates audit and ESG reporting from a single base. At [UniTrust](/en/cases/unitrust), brokers who managed leads and commissions across disconnected channels and unstable spreadsheets moved to a centralized system with real-time visibility of performance. In both, the trustworthy dashboard came after the single source, not instead of it. ## Frequently asked questions **I can't justify investing in fixing the data right now. Where do I start?** Start small and visible. Pick the three numbers that enter leadership's decision and fix only their origin. This is not a year-long data project. It is defining three metrics, reconciling three sources, and proving the dashboard now reconciles. The small case that closes in the meeting finances the next one. **Will the implementation take too much of the team's time?** Less than the current routine already consumes. A good part of the team already spends hours every month exporting from systems, pasting into spreadsheets and checking discrepancies by hand. That time is invisible because it became a habit. Concentrating the effort once, at the single source, gives those hours back instead of repeating them at every close. **How do I measure the result of fixing the data?** With three simple measures, before and after: hours spent per month consolidating the report, the number of times leadership questioned a number in the meeting, and days between month-end and the dashboard being ready. If all three fall, the base improved, and you can show that to whoever approved the investment. ## The right number lives in the base, not the dashboard The question is not which BI tool to buy. It is whether a single source exists where each indicator was defined and reconciled once. Without it, every new dashboard inherits the same divergence with better styling. With it, the choice of tool almost stops mattering. If your dashboard has become a reason for suspicion rather than a basis for decisions, the [Operational Architecture Diagnostic](/en/contact) is a 30-minute conversation to find where the single source is missing in your operation.