As a data analyst I’d measure product-market fit (PMF) using a mix of behavioral metrics and qualitative signals, set practical thresholds, and combine them into a single actionable score and dashboard for stakeholders.
Key metrics/signals (why, threshold, how to track):
- DAU/MAU (engagement): indicates habitual use. Target: DAU/MAU ≥ 20–30% for consumer apps. Track via cohort queries in SQL and visualize trend.
- Retention cohorts (D1, D7, D30): core health metric. Targets: D1 ≥ 35–50%, D7 ≥ 15–25%, D30 ≥ 10–20% depending on category. Compute cohort retention tables and survival curves.
- Activation & funnel conversion: percent of new users completing A→B key action (onboarding → first core use). Target: activation ≥ 40–60%. Instrument events and funnel in analytics.
- NPS / qualitative satisfaction + support volume: NPS ≥ 30 is positive; rising support tickets per user is negative. Combine survey responses, CSAT, and ticket rates.
- LTV:CAC and monetization signals: LTV/CAC > 3 is healthy for paid apps; for ad-driven, ARPU growth and ad-CTR trends matter. Compute using lifetime revenue cohorts.
- Virality / referral rate (K-factor) & organic share: K > 0.2 and ≥50% organic acquisition suggests PMF.
- Usage concentration: fraction of sessions by top X% users — low concentration (more broad usage) is better.
Combining into assessment:
- Build a weighted PMF score (weights aligned with business priorities; e.g., retention 30%, activation 20%, engagement 15%, NPS 15%, LTV/CAC 10%, virality 10%).
- Normalize each metric to 0–100 based on thresholds, compute weighted sum, and define bands (Score ≥75 = strong PMF; 50–75 = product-market fit emerging; <50 = not PMF).
- Complement with statistical tests (cohort comparisons, A/A checks) and trend analyses to avoid noise.
Actionable output:
- Dashboard with score, decomposition, and alerts for falling metrics.
- Recommendations: prioritize fixes (e.g., onboarding if activation low), run experiments, and schedule qualitative interviews when metrics conflict (e.g., high retention but low NPS).
