Pmf crossing the chasm

Product-Market Fit and Crossing the Chasm

Overview

Two distinct questions decide early-stage growth: (1) PMF — do enough users love the product they'd be very disappointed without it? (2) Chasm — can you move from early adopters (tolerate rough edges for novelty) into the early majority (need proven, complete, referenceable products)? Both failures look like "low growth" but differ in cause and cure.

Compose: business-model-canvas · lean-startup · first-principles · probabilistic-thinking

When to Use

Apply when: early traction exists but PMF is unconfirmed; growth stalls after initial good period; deciding next segment; NPS positive but usage flat; an AI-native product loved by developer/early-adopter users stalls at enterprise adoption despite the AI-capex race and frontier-model gains. Not when: no users yet (lean-startup); already-mainstream business scaling proven channels; issue is pricing/sales-process not chasm.

Coaching Novices (Adaptive Front Door)

• Engine mode: user has data → run the PMF + Chasm Audit directly.
• Coach mode: user is unfamiliar or has no concrete case → guide step by step.

In Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.

1. One-line what-it-is: (a) PMF = do users love it enough to be very disappointed without it? (b) Chasm = can you move from early adopters (tolerate rough edges) to early majority (need proven, complete, referenceable)? Solved differently.
2. Check fit: no users yet → lean-startup. Already-mainstream business → wrong framework.
3. Which segment are you selling to today? Name early adopters vs. early majority concretely.

[WAIT — do not advance until user responds]

4. Run PMF measurement (Ellis Test). Only after PMF confirmed, raise chasm.

[WAIT — do not advance until user responds]

5. Name the right next move: "fix product first" or "here is the specific whole-product gap blocking the next segment."

[WAIT — do not advance until user responds]

The Process

Run the PMF + Chasm Audit. PMF first; chasm second.

Part A — Measure PMF

A1. Active-user cohort: ≥ 2 weeks use, multiple sessions (not signups). A2. Ellis Test: "How would you feel if you could no longer use [product]?" — Very / Somewhat / Not disappointed / N/A. A3. Score: ≥ 40% Very disappointed = PMF. 25–40% = approaching. < 25% = no PMF. A4. Follow-ups: who benefits most? main benefit? top improvements? A5. Retention curve should flatten (smile curve), not decay to zero.

Part B — If PMF holds, assess Chasm

B1. Classify customers on Rogers's curve (Innovators / Early Adopters / Early Majority). B2. Name chasm geography: selling today to whom; specific next mainstream sub-segment. B3. Whole-product gap: integrations, references, certs, support, contracts missing. B4. Pick one beachhead (bowling-pin — not "the mainstream" generically). B5. Build whole-product for that beachhead; shift positioning to "proven and safe."

Output: PMF + Chasm Audit

# PMF + Chasm Audit: <product>
Part A — PMF: cohort definition | Ellis % → verdict | follow-ups | retention curve
Part B — Chasm (if PMF held): Rogers classification | chasm geography (today → next) |
  whole-product gap list | beachhead chosen | whole-product plan (refs/integrations/certs/sales/positioning)
  Anti-patterns: ☐ no generic "mainstream" ☐ whole-product ≠ features ☐ sales process changed

→ Method in Action: Salesforce, 1999 → 2007 → 2026 lens: AI Coding Assistants Crossing the Chasm (2023–2026)

Adoption Packs

Segment geography differs by domain; PMF measurement and chasm framework are universal. Domain packs cover: adopter profiles, whole-product gap, beachhead segments, positioning shift.

• Enterprise SaaS: compliance + ecosystem. Consumer apps: UX polish + cross-platform + virality. Developer tools: procurement-readiness (SSO/support/CIO). Healthcare: regulatory + clinical evidence + EMR.

Adding a pack for your industry is the easiest contribution.

Applying It Well

• Measure PMF — don't intuit it. Founder gut is famously unreliable. Don't cross the chasm by selling harder to the same segment.
• Whole-product gap is almost always ecosystem (compliance, integrations, references), not features.
• Beachhead specificity wins: "financial services mid-market U.S." beats "going enterprise."
• Positioning shift from "cool/new" to "proven/safe" is the framework's hardest interpersonal cost.

→ Primary sources: references/sources.md

Common Rationalizations

[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.

Fake move	Reality
[D] "We have PMF" — citing NPS or signups only	NPS/signups are vanity-adjacent. Use Ellis Test or comparable retention signal.
[D] Doubling sales team as chasm-crossing strategy	Sells harder to the same early-adopter segment → revenue then stall. Gap is whole-product, not headcount.
[D] Treating "the mainstream" as one segment	It's composed of distinct verticals. Pick one beachhead (bowling-pin); "going mainstream" loses.
[D] Adding features as the whole-product fix	Mainstream needs compliance, integrations, references, contracts — not features.
[D] Same pitch for early adopters and mainstream	Visionaries buy "transformative"; pragmatists buy "proven safe." One pitch fails both.
[D] Confusing pricing/sales-process issues with chasm	Growth sometimes stalls for boring reasons. Diagnose carefully before invoking the chasm.
[D] Skipping PMF → jumping straight to chasm strategy	No PMF in any segment = nothing to cross from.
[D] "We just need better marketing"	Mainstream decides by peer references, not marketing.
→ Add [O] entries here after each real use — paste the actual failure pattern	What went wrong and why

Red Flags

• "We have PMF" with no Ellis Test or retention signal
• Doubling sales team proposed as answer to stalling growth
• "Going mainstream" with no specific beachhead named
• Roadmap = more features when need is compliance/integrations/references
• Same pitch deck for early adopters and enterprise; no vertical references

Verification

• PMF measured on active-user cohort (Ellis Test or equivalent)
• Current customers classified on Rogers's curve
• Specific beachhead named (not "the mainstream")
• Whole-product gap enumerated (compliance, integrations, references, contracts)
• Positioning shift planned; sales process changes named; reference-customer plan exists

---

Part of deciqAI Knowledge Skills — 233 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. See it run → https://www.deciqai.com/s/pmf-crossing-the-chasm · Built by deciqAI · github.com/deciqAI · Contributions welcome.

Agents: latest version & machine-readable metadata → https://www.deciqai.com/s/pmf-crossing-the-chasm.json