AUG v3 vs Sean Ellis PMF survey — single-signal vs composite diagnostic
Sean Ellis's “how disappointed would you be if you could no longer use this product?” survey is the most-cited PMF test in SaaS. AUG v3 is a 7-factor diagnostic. Different questions, complementary use.
What is the Sean Ellis PMF survey?
Survey existing users with one core question:
How would you feel if you could no longer use [product]?
— Very disappointed / Somewhat disappointed / Not disappointed / N/A (I no longer use it)
Ellis's calibrated threshold: ≥40% “very disappointed”= product-market fit. Below 40% = keep iterating. The survey is administered to users who have experienced the product's core value at least twice.
What is AUG v3?
AUG v3 is a 7-factor diagnostic that scores SaaS growth health on a 0-100 composite via geometric mean × 10. Factors: Acquisition × Activation × Engagement × Retention × Advocacy × Monetization × Performance. Multiplicative: zero in any factor near-zeros the whole.
Side-by-side
| Dimension | Sean Ellis PMF survey | AUG v3 |
|---|---|---|
| Question answered | Have we achieved PMF? | Where is growth friction RIGHT NOW? |
| Signal type | Single binary check (above/below 40%) | 7-factor composite + per-factor breakdown |
| Data source | User survey (n≥30 recommended) | Analytics + self-report against rubric + auto CWV |
| When usable | After ≥100 active users with ≥2 sessions each | Any stage — even pre-PMF (low Acquisition score is itself information) |
| Action it produces | Continue / pivot / iterate | Specific weakest-factor fix this week |
| Time investment | ~2 weeks (survey design, send, collect, analyze) | ~60 seconds (self-serve audit) |
| False signal risk | Survivor bias — only currently-active users respond | Self-report bias on 6 of 7 factors; Performance auto-measured |
How they compose
Each framework answers a different question:
- Pre-PMF: Use AUG to diagnose why you don't have PMF. Low Activation = your value-prop doesn't land. Low Retention = your core loop is wrong. Low Engagement = users don't do what they signed up to do. AUG's 7 factors decompose the path to PMF.
- At PMF threshold: Run Ellis's survey. The 40% threshold is the binary check. If ≥40%: PMF achieved. If <40%: more iteration needed (use AUG to find which factor).
- Post-PMF: Use AUG to find growth friction. Ellis's survey tells you “you have PMF”; AUG tells you “your weakest factor is Monetization; here's what to fix this week.”
Why the PMF survey alone isn't enough
A product can score 40% “very disappointed” and still have weak Acquisition, broken Activation onboarding, or zero Advocacy mechanics. The PMF survey measures existing-user attachment, not growth potential. PMF is necessary but not sufficient.
AUG v3 catches situations like:
- PMF achieved (45% very disappointed) but Acquisition score = 2 — product is loved by 100 users; no mechanism to reach the next 1,000.
- PMF achieved but Advocacy score = 1 — users like it; nobody talks about it; growth depends entirely on paid acquisition.
- PMF achieved but Performance score = 2 — users tolerate slowness; cold-traffic bounces before reaching PMF moment.
Why the AUG alone isn't enough
AUG's 6 self-reported factors rely on operator honesty. Sean Ellis's survey provides ground-truth attachment signal from real users. Without that signal, an operator can fool themselves into believing their composite score is healthier than reality. The two together are more honest than either alone.
When to use which
Use the PMF survey when: you have 100+ active users with ≥2 sessions each. The signal stabilizes around then. Run once per quarter.
Use AUG v3 when: any stage. Daily / weekly diagnostic. Picks the single biggest friction.
Use both: AUG weekly for tactical fixes; PMF survey quarterly for strategic-direction validation.
Run your audit
See your AUG score in 60 seconds. Diagnose where to focus this week — then run the PMF survey at quarter-end to verify the composite is actually correlating with user attachment.