← Docs
Recipe

PMF survey analyzer

Run the Sean Ellis PMF survey, pipe results through Meridian, and get a scored segmentation in under 60 seconds.

Ingredients

  • One CSV export from Typeform / Google Forms
  • The canonical question: “How disappointed would you be if you could no longer use [product]?”
  • Meridian API key with survey:read scope

Steps

  1. 1Upload CSV to /api/v1/surveys/import
  2. 2Meridian auto-detects the disappointment column and buckets responses: Very, Somewhat, Not at all.
  3. 3Call POST /api/v1/surveys/analyze with method=pmf
  4. 4Read the response: overall PMF score, persona clusters, and verbatim highlights from the “very disappointed” segment.

Output shape

{
  "pmf_score": 0.42,
  "segments": [
    { "bucket": "very_disappointed", "count": 84, "pct": 42 },
    { "bucket": "somewhat", "count": 72, "pct": 36 },
    { "bucket": "not_at_all", "count": 44, "pct": 22 }
  ],
  "personas": [
    { "label": "Power users", "size": 84, "top_quote": "..." }
  ],
  "threshold_met": true
}

Interpretation

≥40%
Strong PMF signal
25–39%
Niche traction
<25%
Revisit positioning