Pretend you are the growth PM so funnel gaps, experiment theater, and metric blur fail before a growth package absorbs them.
Optimistic growth packages optimize for "activation will rise." The growth PM seat does the opposite. It asks which experiment invents lift the instrumentation will not support, which funnel step dual-counts the same user as activated and as retained, which "required" holdout is already optional in practice, and which cohort still lacks a named owner. A neat funnel chart is not evidence that next quarter's growth will clear.
Mandate the seat
Name a real job: underwrite an experiment roadmap without inventing engineering capacity, clear an activation claim that analytics can reconcile under load, or defend a funnel rewrite without dual-counted conversion rates. Give constraints: the evidence standard for experiment wins, the cohorts you will refuse to leave unowned, and the lift claims you will not teach when instrumentation is not ready. Without constraints the seat becomes cartoonish. With constraints it produces questions you might actually hear in a growth review, a board prep, or a fight over who owns onboarding drop-offs.
Prompt example: "Growth PM: list the top reasons to delay this growth package, the funnel step with the weakest owner, the lift claim that worries you most, and the ten diligence questions you would send after review. Stay inside a realistic mandate."
Outputs worth keeping
- Top reasons to challenge, delay, or rewrite this growth package.
- The activation or retention claim that looks strongest and is least evidenced.
- The cohort, channel, or surface that would break first under a traffic spike.
- What would make you accept residual experiment risk in writing.
- The ten hardest follow-up questions after the meeting.
Run that brief in Pingpong against the real experiment backlog, instrumentation map, holdout notes, and open risk list. Follow with a home-team response pass so you leave with edits and source packs. When the decision is a public launch claim, run this seat after analytics and product attacks so it can use earlier objections as ammunition.
When the plan leans on a single viral loop, a single "we'll A/B it later" promise, or a single hero channel, force the seat to price concentration risk in writing. Ask what happens if the event taxonomy still hosts retired names, if holdouts still lack a named steward, or if marketing keeps teaching lift language product already retired. Pair with pretend you are the demand gen lead, pretend you are the CPO, war-game a growth experiment, war-game an experimentation policy, and the war-game decisions hub. See how to run a Pingpong.
If the package coincides with a pricing or packaging change, ask the growth PM seat to map every claim that still assumes last quarter's funnel definitions. Landing pages, emails, and in-product prompts count. A growth memo that looks clean in a slide while analytics still posts retired event names will fail on the first experiment read.