Semantic layers: one revenue number for the board

September 25, 2026 · 5 min · Free, with full transcript

What happens when finance, sales and product each compute revenue their own way? The episode opens on a board meeting where three presenters defend three different quarterly numbers, and argues that the cost is not bad data but frozen decisions while teams re-litigate whose spreadsheet counts. Finance recognizes revenue on fulfillment, sales on signature, product on ship date, and each team's bonus depends on its version. You come away with a build sequence for a semantic layer: inventory the 12 to 15 metrics that appear in board decks and budget fights, lock each to one written formula with finance and sales signing off together, then choose tooling. The discussion weighs a dbt Labs vendor claim against MIT Sloan Management Review's work on trust in shared numbers, and covers churn, customer acquisition cost, and the CDO's refereeing role.

Listen to the podcast

4 min

Chapters

Key takeaways

  • Inventory only the 12 to 15 metrics that show up in board decks and budget fights, not all 400.
  • Lock each metric to a single written formula and make finance and sales leads sign off in the same room.
  • Define the five metrics in your next board meeting, ship those, and use the momentum for the rest.
  • Treat the dbt Labs figure on conflicting metric definitions as vendor marketing, not evidence.
  • Email each department head before the next board meeting asking for their exact formula in one sentence, then compare the replies.
Read the full transcript

Host:Leaders' Insights, CDO Edition. We get into Airbus lost a quarter to misaligned revenue metrics. Here is the playbook that prevents it. Picture a Tuesday board meeting where three people present three different revenue numbers for the same quarter, and all three swear they're right.

Expert:That happened somewhere last month. It happens more than anyone admits.

Host:And this is the thing that just cost Airbus a full quarter.

Expert:Effectively, yes. Finance, sales, and product each had their own definition of revenue, and nobody noticed until the numbers collided in a board deck. Decisions froze while everyone re-litigated whose spreadsheet was the real one.

Host:How does a company that builds flying machines to the millimeter not agree on what a sale is?

Expert:Because measuring a wing is physics and measuring revenue is politics. Finance recognizes revenue when the contract's fulfilled. Sales books it the day the ink dries. Product counts it when the feature ships. All defensible, all different, and each team's bonus quietly depends on their version winning.

Host:So when the CFO says revenue grew 8 percent...

Expert:Someone in sales is thinking 12, and someone in product is thinking 4. Nobody's lying. They're just speaking three dialects of the same word.

Host:Fine. Give me the fix and start from zero, because I'm the exec who's never heard the phrase.

Expert:The fix is called a semantic layer. Think of it as a single dictionary sitting between your raw data and every dashboard, chart, and deck. Instead of each team writing its own formula for revenue, the definition lives in one place, and everyone pulls from it.

Host:A dictionary? That's the whole revolution?

Expert:The revolution is that it's enforced. Right now, active customer means whatever the analyst who built the report decided at 11 p.m. A semantic layer makes the definition a shared organizational fact. You can't quietly redefine it in your corner. Change it once, it changes everywhere, and there's a record of who changed it.

Host:Walk me through building it. Step one.

Expert:Step one. You inventory the metrics that actually drive decisions. Not all 400. The 12 or 15 that show up in board decks and budget fights. Revenue, churn, the share of customers who walk each month, customer acquisition cost, that sort of thing. Step two. You lock each one down to a single written formula, and you make the finance and sales leads sign off in the same room. That meeting is ugly. It's supposed to be. You're surfacing a disagreement that was already costing you money silently.

Host:Step three, presumably, is the software.

Expert:Right, and this is where I'll be careful. DBT Labs, who sell exactly this kind of tooling, say organizations run somewhere north of a dozen conflicting definitions of core metrics before they consolidate. Directionally, that matches what I see. But they're a vendor with a product to move, so treat the number as a sales flyer, not gospel.

Host:Is there anyone neutral backing this up?

Expert:MIT Sloan Management Review has written for years that data-driven firms outperform mainly when everyone trusts the same numbers. Trust being the operative word. The tooling is plumbing. The trust is the point. A pipe that delivers water nobody drinks is just an expensive noise.

Host:Here's the uncomfortable one. This sounds like a project that takes 18 months and dies in committee.

Expert:It dies in committee when you try to define all 400 metrics by consensus. Don't. Pick the five that appear in the next board meeting, define those, ship them, and let people feel the relief of arguing about strategy instead of arithmetic. Momentum buys you the rest.

Host:And the CDO's real job in this?

Expert:Refereeing, mostly. The technology is solved. The hard part is telling the VP of sales that his favorite number is now computed the finance way and surviving the conversation. If you're not making one senior person mildly annoyed, you haven't actually centralized anything. You've just built a fourth definition and called it a truce.

Host:Give me the one thing to do tomorrow.

Expert:Before your next board meeting, take your top five metrics, email each department head, and ask them to write down their exact formula in one sentence. Compare the replies. The gaps you find are the quarter you're about to lose. And now you'll see it coming.

Host:This episode draws on DBT Labs, vendor, data tooling, MIT Sloan Management Review. That's it from us. The reading continues at MBA-training.com, comma, new CDO analysis every day.

Read and go deeper