First: don't pick a side. The instinct 'my data is right, theirs is wrong' kills the investigation.

The reconciliation process:

Define 'revenue' side by side. Finance ledger revenue is usually 'recognized revenue net of refunds and taxes.' Analytics 'revenue' is often 'gross transaction value before adjustments.' These should be different — the question is by how much, not which is right.
Check the time boundary. Finance uses fiscal close dates (often month-end-plus-a-few-days for booking); analytics dashboards tend to use calendar midnight. Adjustments finance made this week apply to last week's data.
Check segments. Are both definitions including the same subsidiaries / product lines / geos? A recent acquisition or divestiture changes the answer.
Check for canceled / refunded transactions. Analytics might count them on the transaction date; finance might net them on the refund date.
Check currency conversion. Analytics often uses average exchange rates; finance uses point-in-time.
Check taxes / shipping / tips. Which are in / out of each definition?

A likely finding:

"Analytics 'revenue' is GMV — total transaction value, before refunds, taxes, and net of promotions. Finance 'revenue' is recognized revenue under ASC 606 — excludes pending refunds, deducts taxes and shipping, uses month-end cutoffs that move Q3 revenue into Q4 adjustments. The 7pp gap (5% up vs. 2% down) is almost entirely explained by: (a) refund recognition timing (-3pp), (b) Q3 revenue recognition lag (-2pp), (c) promotional spend netting (-2pp). Both definitions are internally consistent; they measure different things."

What to do about it:

Document the definitions side by side. Put them in a shared spec so every future conversation is about the same numbers.
Build a reconciliation dashboard. A view that shows analytics revenue, finance revenue, and the line-item deltas between them. When they diverge in the future, the dashboard tells you why.
Align on which to cite externally. For earnings, use finance. For product decisions, use analytics. Don't mix in the same meeting.

What NOT to do:

Argue to the CFO that your number is right. Finance owns the definitional authority on revenue — pick your battles.
Adjust your dashboard to match finance without investigating. You'll just have one confused number instead of two conflicting ones.

Stakeholder framing:

"Finance's number is correct for earnings reporting; our number is correct for understanding gross product performance. They're answering different questions. Let me send you a side-by-side with the bridge, and we can decide which to cite in which context."

What senior candidates add:

This reconciliation work is often the most valuable thing a senior analyst does — it's the bridge between business reporting and product analytics. Companies without one suffer repeated number disputes.
Make the bridge durable: automated nightly reconciliation that alerts when the delta exceeds a tolerance band.

Two metrics disagree. How do you reconcile?

Draft your answer

Stuck? Peek at a hint

More product metrics questions