Laws of PM Discovery Segel's Law

Segel's Law

Two Watches, No Clarity

Having multiple conflicting data sources doesn't give you more truth — it gives you more doubt. Every question needs one source you agree to trust.

Why PMs should care

Having multiple sources of truth — Amplitude, a data warehouse, a spreadsheet the product team maintains, Finance's revenue dashboard — doesn't give you more truth. It gives you more doubt. Every meeting stalls on 'which number is right?' instead of 'what should we do about the number?'

The fix isn't to pick a single tool (usually not possible anyway). It's to pick a single source per question: Amplitude for product behaviour, the warehouse for revenue, Finance for regulatory reporting, one specific spreadsheet for cohort value.

When sources disagree, reconcile it in writing and in public. A Confluence page called 'why Amplitude and Finance disagree on sign-ups' is worth more than a thousand Slack messages. The next time the question comes up, you link to the page instead of having the argument again.

Example in product work

Amplitude says 12,400 sign-ups last week. The Finance dashboard says 11,980. The product-team spreadsheet says 12,150. In the CFO's weekly review, the answer 'around twelve thousand' is not acceptable.

The PM spends an afternoon reconciling and finds: Amplitude counts sign-up events, Finance counts completed KYC, the spreadsheet pulls from a cached report that misses the Friday evening backfill. All three are correct — they measure different things.

The output isn't a fix; it's a definition page. From then on, 'sign-ups' in the roadmap means Amplitude's number, 'active customers' means Finance's, and nobody uses the spreadsheet for the CFO again. Same data, dramatically fewer meetings.

What to do when you see it

Sources & further reading

Back to all 59 laws