A Recommendation Should Come With an Expiry Condition
Teams document the recommendation and the numbers behind it. The assumption carrying it usually stays in the meeting — so nothing reopens the decision when the evidence changes.
A recommendation should come with an expiry condition.
Not necessarily a date. It should be clear what change in the evidence, assumptions, or business context would make the recommendation deserve another look.
Most operating recommendations are made using what the team knows at a particular point in time. A team may increase interstitial frequency because ad ARPDAU has room to move, scale a campaign because the early cohorts are recovering well, or extend a LiveOps event because participation and revenue remain strong.
Each may be the right decision. But each is also conditional.
The increase in ads assumes that the additional revenue will be worth any movement in retention, payer behaviour, or longer-term LTV. The decision to scale assumes that the cohorts are mature enough, the product can absorb more volume, and the next set of users will behave broadly like the ones already measured. The event extension assumes that current engagement is not creating fatigue, cannibalising the next event, or weakening what comes after it.
The assumption stays in the meeting
Teams are generally good at documenting the recommendation and the numbers supporting it. The assumption carrying the recommendation is often left inside the meeting.
That becomes a problem when the context changes. Retention may weaken, cohort quality may shift, or participation may start declining. The dashboard can show all of it, but the original decision is not automatically reopened because nobody defined which movement should bring it back into review.
The recommendation slowly becomes the way the team operates, even though the evidence behind it is no longer the same.
Four things to make explicit
- What evidence currently supports it.
- What assumption needs to remain true.
- What signal should trigger another review.
- Who owns that review.
The last point matters. Otherwise, the trigger becomes another alert that everyone assumes somebody else is watching.
This does not mean teams should keep reopening every decision. It means temporary evidence should not quietly turn into a permanent operating rule.
A recommendation is incomplete if it explains what the team should do, but not what would make the team reconsider it.