Reporting and KPIs

Different departments report different numbers?

When different departments report different numbers for the same metric, leadership reviews stall on whose figure is right. Here's the fix.

Austin Levine

·

Aktualisiert

When different departments report different numbers for the same metric, it's usually because each team defined that metric on its own. Sales counts revenue at the signature, finance at the invoice, leadership at recognition. One written definition per number, built into one shared model, ends the disagreement.

When different departments report different numbers for the same metric, the damage shows up in a room. You're in a leadership review, slide three says revenue was 4.2 million, slide nine says 4.5 million for the same quarter, and the next twenty minutes go to arguing about which slide to believe. Nobody decides anything. The meeting was supposed to be about what to do with the quarter, and instead it's about whose spreadsheet wins.

This is one of the most common reasons leaders stop trusting their own reporting. Only 40% of executives say they always trust their data (Talend). When two numbers disagree in front of leadership, that distrust stops being a survey stat and becomes the tone of the meeting.

The frustrating part is that both numbers are often correct. Sales isn't wrong. Finance isn't wrong. They're measuring two different things and calling them the same word. Let's look at why that happens, what it costs, and the decision that fixes it.

Key takeaways

  • Departments report different numbers because each team defined the same metric on its own.

  • Sales counts revenue at signature, finance at invoice, leadership at recognition.

  • One written definition per number, built into one shared model, ends the disagreement.

  • Fix the handful of headline numbers first, since most disagreement comes from those.

Why do different departments report different numbers for the same metric?

Take the word "revenue." It sounds like one thing. In practice, three teams count it at three different moments.

Team

What they call "revenue"

When they count it

Sales

Signed deal value

The day the contract is signed

Finance

Invoiced amount

The day the invoice goes out

The leadership

Recognized revenue

The month the service is delivered

All three are valid ways to see the business. A sales leader wants to know what the team closed. Finance wants to know what was billed. The leadership wants the number the auditors will accept. The problem starts when all three build a report, label the top number "Revenue," and nobody wrote down which one they meant.

It gets worse one layer down. Each report usually sits on its own copy of the data, with its own hand-built formula. Five people build five slightly different versions of the same measure, each on a spreadsheet or a private dataset. 94% of business spreadsheets used in decisions contain errors (Panko et al., 2024), so every extra copy is another place for the number to drift. This is the plain reason reports show different numbers: they were never built from one shared definition, so they were never going to match.

Data silos make it stick. 82% of organizations say silos disrupt their workflows (ZoomInfo). When each department owns its own tools and its own version of the truth, there's no shared place where "revenue" is defined once and reused everywhere.

What's it costing you when the numbers disagree?

The cost isn't only the awkward meeting. It compounds in three ways your finance and data teams feel every week.

Decisions get made on the wrong number. If leadership sizes a hiring plan off the sales figure when it should have used recognized revenue, the plan is built on air. You don't find out until a quarter later.

The month-end close drags. Teams spend the first week of every month reconciling figures by hand instead of reporting them. A healthy close lands in about eight days, and the strongest finance teams close in under three (SolveXia). If yours runs longer, a chunk of that time is people chasing down why two reports don't agree.

Your reporting tools go unused. When people can't trust the dashboard, they export everything back to a spreadsheet and check it by hand. Around 70% of dashboard work ends up wasted this way (Reveal BI). You paid for Power BI, and your teams still live in Excel, because a number they can't defend is worse than no number at all.

Poor data quality carries a hard price on top of all this: Gartner puts the average cost at 12.9 million dollars a year. A lot of that's the quiet tax of everyone maintaining their own version of the same figures.

What does good look like when everyone trusts one number?

Good is boring, and that's the point. In a company that has this sorted, a leader opens a report and doesn't wonder where the number came from. There's one figure for revenue, one owner who can answer for it, and a small note on the page showing the data is current as of last night.

Three things are true when it works:

  • There's one agreed definition per metric. Revenue means one specific thing, written in one sentence, and every report uses it. When someone asks "which revenue," there's an answer on file.

  • There's one source, not five. Each number is built once, in a shared model everyone reports from, so sales and finance are reading from the same place. If it's wrong, it's wrong for everyone at once, which means it gets caught and fixed fast.

  • People can see it's current. The report shows when it last refreshed, so nobody wonders if they're looking at last week's figures.

That's what a single source of truth actually means in day-to-day terms. Not a grand platform. One definition, one build, visible freshness. The teams that reach it didn't buy a fancier tool. They agreed what the words mean and then kept everyone to it.

How do you get to one agreed definition?

You don't need a year-long program to start. Most of the disagreement comes from a handful of headline numbers, so you fix those first. Here's the decision-level sequence.

Pick the numbers that show up in the management report. Usually revenue, margin, pipeline, headcount, and cash. Ten at most. These are the ones worth arguing about, so these are the ones you settle first.

Write one sentence per number. For each, agree what it counts, where it comes from, and how often it updates. "Revenue is recognized revenue, from the finance ledger, refreshed nightly." That one line is the contract. Sales, finance, and leadership sign off on it once, and the debate is over.

Point every report at one source. Once the definition is agreed, each number should be built once, in a shared model, and reused. No more private copies of the revenue formula living on individual laptops.

Give each number an owner. One named person, not a team alias, who can answer "why is this figure what it is." When nobody owns a number, nobody fixes it when it looks off.

Ask your team for those four things and you've covered most of the gap. The written definitions are the part people skip, and they're the part that ends the arguments.

What your team actually does about it

One level deeper, for the person who'll own the build: this lives in the semantic model. Your data team defines each measure once, in a certified shared dataset, so every report inherits the same revenue calculation instead of re-inventing it. Access gets set by role with row-level security, so a regional manager sees only their region, off the same trusted model. And a scheduled refresh with a visible "data as of" stamp closes the loop, so people can see the figure is current. This is a few weeks of focused work on your top reports, not a rebuild of everything. That's the layer that turns an agreed definition on paper into a number leadership can't argue with.

When should you bring in help?

Some teams settle the definitions in a couple of workshops and handle the build in-house. Bring in outside help when the disagreements keep coming back after you've tried to fix them, when the close is stuck above a week, or when the same number still shows up three different ways no matter how many meetings you run. Those are signs the problem is in the model underneath, not in the reports on top, and that's the part that needs someone who does it full time.

Frequently asked questions

Why do reports show different numbers for the same thing? Because each report was built from its own copy of the data with its own version of the formula, and the metric was never defined in one shared place. Two teams can both be correct and still disagree, because they're counting the same word at different moments.

Isn't this just a tooling problem we can buy our way out of? No. A better tool with no agreed definition just produces prettier versions of the same conflicting numbers. The fix is a written definition per metric plus one shared model to build it from. The tool matters after that, not before.

What's a single source of truth, in plain terms? It's one agreed definition for each number, built once in a shared model, that every report reads from. So when anyone asks for revenue, they get the same figure whether they open the sales report or the finance report.

How long does it take to fix? Agreeing the definitions for your top ten numbers takes a couple of focused sessions. Building them into one model usually takes a few weeks for the reports that matter most, not a year. You feel the difference at the first leadership review where nobody argues about the figure.

Who should own the definition of a metric? One named person per number, usually in finance for financial metrics, who can answer for it. A team alias doesn't work, because when everyone owns it, nobody does.

Where to start

If you want to see where your reporting stands before you commit to anything, run the free Power BI Governance and Cost Scorecard. It's a five-minute check that scores where your numbers can drift and shows the top risks to fix first, with no email gymnastics.

And if your numbers still won't reconcile after you've agreed the definitions, that's the work we do at CaseWhen. You can book a semantic-model and AI-readiness review with us, a plain look at why the figures disagree and what it takes to get to one number everyone trusts.

Read next:

  • Data governance framework: why most of them get ignored, the pillar on owners, one certified model, and access by role.

  • KPI dashboard in Power BI executives actually trust, the build side, from a blank file to a leadership team-ready page.

Ist Ihr Power BI richtig aufgesetzt?

1

Buchen Sie 45 Minuten und sagen Sie uns, worum es geht.

2

Wir gehen Ihre Berichte und Ihre Umgebung live mit Ihnen durch.

3

Innerhalb von zwei Arbeitstagen erhalten Sie unseren Plan schriftlich.

Ist Ihr Power BI richtig aufgesetzt?

1

Buchen Sie 45 Minuten und sagen Sie uns, worum es geht.

2

Wir gehen Ihre Berichte und Ihre Umgebung live mit Ihnen durch.

3

Innerhalb von zwei Arbeitstagen erhalten Sie unseren Plan schriftlich.

Ist Ihr Power BI richtig aufgesetzt?

1

Buchen Sie 45 Minuten und sagen Sie uns, worum es geht.

2

Wir gehen Ihre Berichte und Ihre Umgebung live mit Ihnen durch.

3

Innerhalb von zwei Arbeitstagen erhalten Sie unseren Plan schriftlich.

CaseWhen ist eine BI-Beratung aus Berlin. Wir bauen Reporting, dem die Geschäftsführung vertraut, auf dem Microsoft-Stack: Power BI, Fabric und Azure.

Berlin, Deutschland

© CaseWhen Consulting GmbH

Deutsch

CaseWhen ist eine BI-Beratung aus Berlin. Wir bauen Reporting, dem die Geschäftsführung vertraut, auf dem Microsoft-Stack: Power BI, Fabric und Azure.

Berlin, Deutschland

© CaseWhen Consulting GmbH

Deutsch

CaseWhen ist eine BI-Beratung aus Berlin. Wir bauen Reporting, dem die Geschäftsführung vertraut, auf dem Microsoft-Stack: Power BI, Fabric und Azure.

Berlin, Deutschland

© CaseWhen Consulting GmbH

Deutsch