| name | funnel-stage-lift-needs-downstream-capacity-check |
| description | Guard the step where a VALID early-funnel lift becomes a recommendation. A segment genuinely
starts/converts more at stage N — and the write-up is about to say "prioritize this segment"
as if that buys the END outcome (enrollment, purchase, activation). Use when: (1) a finding,
headline, card, or play recommends tilting effort toward a segment based on a stage-N rate
(lead→start, start→submit, signup→trial); (2) the favored segment funnels into a
capacity-capped or selective downstream stage (limited seats/inventory/approval slots, an
admissions committee, a review queue); (3) you are REWORDING or promoting an existing
funnel-stage finding — the reword inherits the original's stage scope, so re-check it;
(4) a stakeholder asks "but doesn't that program/product only take a few?". The lift can be
real AND the recommendation still over-promise: extra stage-N entrants feed a bottleneck,
not the end outcome. Distinct from cohort-milestone-lift-is-funnel-position-not-effect
(which attacks the lift's VALIDITY); here the lift is valid — the question is what decision
it licenses. Fix = check downstream capacity + segment-specific downstream yield, then state
the claim's stage ON the artifact face, not in a collapsed footnote.
|
| author | Claude Code |
| version | 1.0.0 |
| date | "2026-07-10T00:00:00.000Z" |
| disable-model-invocation | true |
A valid stage-N lift is not an end-outcome claim — check downstream capacity before recommending
Problem
A segment shows a real, verified lift at an early funnel stage — say, prospects who name an
interest in a popular program area start applications at ~1.7× the base rate. The natural
write-up is "prioritize this segment." But to the reader, "prioritize them" silently means
"this buys more of the END outcome" (enrollments, purchases). If the favored segment funnels
into a supply-constrained downstream stage — selective programs admitting ~1 in 10 decided
applicants while open programs admit ~19 in 20, limited inventory, an approval quota — the
extra stage-N volume hits a wall the claim never mentioned. The finding is true; the implied
promise is not.
Context / Trigger conditions
- Any recommendation of the form "tilt/lean/prioritize toward segment X" grounded in a
lead→start, start→submit, or other early-stage rate.
- The favored segment concentrates in capacity-capped or selective downstream units.
- A REWORD or promotion pass on an existing funnel finding — rewording inherits the original
stage scope, so the check must be re-run, not assumed.
- A stakeholder reads the claim and immediately asks about seats/capacity — treat that as the
canary: if one reviewer went straight there, so will the audience.
- Trigger phrases: "these leads convert best, focus on them", "grow the X pipeline",
"prioritize the high-interest segment", "this segment is our best top-of-funnel bet".
Solution — the two-question downstream check (run BEFORE the framing ships)
- Supply constraint: is the segment's destination capacity-capped or selective? Pull the
downstream acceptance/allocation rate for the favored segment's units vs the open ones
(e.g. accepts-per-100-decided by program). If the favored units sit far below the open
ones, the stage-N lift buys starts, not end outcomes.
- Segment-specific downstream yield: does the segment's accepted→committed /
committed→completed conversion differ from baseline? Compute it per unit — and expect the
bottleneck LOCATION to move: in one real case, two selective units lost ~2/3 of their
admits after acceptance (high accept-side throughput, low yield), while two others were
brutally selective at the accept step but their admits committed above average.
"Selective ⇒ lower downstream yield" is an assumption, not a fact — check, don't infer.
- State the claim's stage on the artifact face. If the caveat lives only in a collapsed
drawer/appendix, a scanning reader reads the stage-N claim as an end-outcome claim. One
line at the recommendation ("this grows application starts, not seats — the favored
programs are capped; the yield side lives in the accepted→committed analysis") is the fix.
Link the downstream analysis rather than duplicating its numbers (construct drift risk).
- Keep the lift and the promise separate in the action line. The defensible action is
stage-scoped ("use the interest signal as a priority marker for completion nurture"),
not outcome-scoped ("this segment will drive enrollment").
Verification
- The artifact face names the stage the claim lives at, adjacent to the recommendation.
- The downstream capacity/yield numbers cited come from a committed, re-runnable derivation
(per data-provenance-verifier), not from memory or assumption.
- A cold reader (or a fresh-reader review pass) can answer "does this buy end outcomes?"
without opening any drawer.
Example
A dossier card recommended leaning outreach toward prospects naming a high-demand program
area (real stage-1 lift: ~1.7× application-start rate, verified). The downstream check found
the area's flagship program admits ~11 of 100 decided applicants (vs ~95 for open-admission
programs) — and, per-program, the post-acceptance yield SPLIT: two programs leaked ~2/3 of
admits after acceptance while two equally selective ones converted admits above average. The
shipped fix was one line on the card face — "this grows application starts, not seats" — plus
a pointer to the accepted→committed card for the yield half. The recommendation survived; the
over-promise did not.
See also
cohort-milestone-lift-is-funnel-position-not-effect — when the lift itself is suspect
(composition/funnel position); run that first if the lift isn't yet verified.conditional-funnel-by-segment-within-level — locating WHERE in the funnel a segment acts.differentiator-scoping-by-provenance-not-signal — marker-vs-lever framing discipline for
the surviving action line.ship-the-correction-to-every-rendered-surface — the added caveat must reach every surface
that renders the recommendation (page, export, design copy).
