Moore Positioning Findings - AFPS Tracker

The framework-selection gate was approved earlier (compiled YAML, approval_status: ready-for-agent-review) — that is the Stage-1 scope approval for the whole selected set, so moore-positioning enters directly at its research stage. Frameworks 1 (jtbd-positioning) and 2 (strategic-canvas) are complete and their intermediates are written; this session ran the final pending framework, moore-positioning, inline.

The Geoffrey Moore “Crossing the Chasm” template, filled from approved evidence. Each blank is graded for evidence quality in the next section.

Compressed claim: AFPS Tracker is the passive, source-of-truth-preserving way to stay in command of a multi-branch AI research portfolio — navigable like a board, faithful like the files.

Confidence per template element. The chart shows relative evidence strength; the full status + source table follows (open with “View as table”).

Status key: Evidenced = directly supported by approved research artifacts; Hypothesized = reasonable inference, not directly confirmed; Unvalidated = assertion without supporting evidence.

Weakest element: Market category

• Why it’s weak. Moore’s template treats category as the frame buyers use to find and evaluate a product. Here there is no established category and no search term: the canvas concluded “no existing category cleanly fits,” competitive analysis found no validated public AFPS-native competitor, and Stage-0 interrogation could only offer a provisional internal label (“AFPS workflow-state tracker”). A position whose category is invented is hard to discover and easy to mis-evaluate.
• What would change if wrong. If buyers reach for an existing frame (generic PM, research repository, agent memory) the product gets compared on that frame’s terms — breadth, collaboration, execution — exactly the red-ocean axes where the deliberate “Low” scores read as missing features instead of discipline. Mis-framing turns the wedge into a weakness.
• Evidence needed. Discovery interviews on the words operators actually use for this pain; whether they expect a “tool” at all or treat it as a native-files problem; whether a borrowed anchor (“portfolio tracker for AI research workflows”) aids comprehension without inviting PM expectations.
• Cost of being wrong. High for go-to-market (invisible or mis-shelved); moderate for the product itself (the differentiator still holds even if the label changes).

Second weakest: existence + urgency of the target/need at scale

Not a template-wording problem but the deepest risk. The target customer and need are well-evidenced from artifacts but not user-tested: whether AFPS operators exist beyond dogfooding, and whether multi-path pain clears the bar to fire free native files, remains the existential open question (icp.md, journey-map, jtbd all flag it; WTP unproven).

Category-label alternatives considered (within this element)

Alternative 1: Studio / consultant decision-provenance frame

Why considered: strongest adjacent WTP (business budget, billable-time logic); provenance is a real social facet of the job. Why not preferred: studios are the adjacent ICP, not the approved primary; AFPS adoption among them is unproven; it over-weights the provenance/social facet jtbd explicitly marks secondary, pulling toward a client-reporting tool and away from the individual-operator wedge.

Alternative 2: Agent re-grounding / persistent-memory frame

Why considered: matches operator “context loss / keeping up with AI speed” language; the agent-re-grounding job is a genuine functional facet. Why not preferred: reframes the product as agent memory / control plane (Runshift territory), which jtbd and competitive analysis place out of approved scope; it inverts the human-legibility anchor and trades a clear wedge for a crowded execution-adjacent one. Keep as latent expansion, not the headline.

• No direct AFPS-operator interviews or observed switching sessions; the template is synthesized from approved artifacts, not live customer language.
• No measured reconstruction-time baseline, so the size of the need (and the benefit) is directional.
• No prototype test of the source-linked read-through (the predicted aha that underwrites benefit + differentiation).
• No direct WTP evidence; adjacent pricing is a budget signal only.
• Category framing is untested with real buyers — the single biggest template risk.
• No parser-fidelity audit; differentiation defensibility assumes reliable read-only parsing.

Cross-framework note (for parent synthesis, not routing): this is the final selected framework (3 of 3). The Moore template lands in a new/taught category, corroborating strategic-canvas’s “no category fits” finding and strengthening the case for synthesis to weigh whether to revisit category-design. This note does not itself invoke any skill.

On approval (final compiled YAML, approval_status: ready-for-agent-review), the parent loop will:

No canonical research/afps-tracker/positioning.md is written by this framework. With all three intermediates present, the next run enters synthesis (State B) and builds the unified positioning.md review page.

Review Gates

Each gate below requires one answer. “Other / None of the above” and “Need clarification” are always available; selecting them (or leaving a gate unanswered) keeps the response in not-approved.