Why your reports disagree, and why a new dashboard will not fix it
When two departments bring different numbers to the same meeting, the problem is almost never the reporting tool. It is that nobody has decided what the number means.
Written for chief executives, finance directors and heads of operations.
The meeting everyone recognises
Finance says the figure is one thing. Operations says it is another. Both are reading from a system that is working exactly as designed, and both are right within their own definition. The meeting stops being about the decision and becomes about whose number to trust.
The instinct at that point is to buy a reporting tool, because the symptom looks like reporting. It rarely is. A dashboard drawn over two systems that disagree produces a faster, prettier disagreement.
Three causes, none of them the dashboard
In the work we have done, disagreement between reports almost always comes from one of three places, and each has a different fix.
- The same entity exists more than once. One customer is three records with three identifiers, so any count of customers depends on which system you asked.
- The same word means different things. Active, complete, churned and overdue each carry a definition that was set inside one system for one purpose, and was never agreed across the organisation.
- The numbers are assembled by hand. Exports are stitched together in a spreadsheet, which means the figure depends on who did it, when, and which export they used.
What actually fixes it
Resolve identity first. Decide what makes two records the same customer, apply it consistently, keep the history of what was merged, and handle the conflicts rather than discarding them. Until this is settled, every downstream number is an estimate.
Then define each metric once, in one place, and have every report read that definition rather than recalculating it. This is the part organisations skip, because it is a series of uncomfortable conversations rather than a piece of software. It is also the part that ends the argument permanently.
Only then is the dashboard worth building, and by that point it is the easy half. On one integration and reporting engagement, the management pack stopped being a monthly project and became a live link, and the discussion moved from whose number was right to what to do about it. The tooling was not the interesting part of that work.
You probably do not need to replace anything
The systems holding the truth are usually doing their jobs. What is missing is the layer between them: translation, synchronisation and an agreed model, so each system keeps its own shape while the organisation sees one operation. Replacement is expensive, slow and risky, and it should have to earn a business case rather than a preference.