Product Metrics and Cohort Retention
Define KPIs with exact specs and read cohort, retention, and adoption data right: immature cohorts, mix shift, and real denominators.
Before & after
Same prompt, same agent — the only difference is the skill.
Good question. You'd want to set up some kind of analytics to track how people are moving through the new flow. You could look at completion rates, time to completion, drop-off points, that kind of thing. Retention might be a good leading indicator—if people complete onboarding, do they come back? You could compare the new flow against the old flow and see if retention is higher. You might also want to look at feature adoption after onboarding. The key is defining what 'working' means for you. Is it faster completion? Higher retention? More feature adoption? Once you nail down your KPI, you can set up tracking in your analytics tool and measure it. Do you have any user feedback yet? That might also tell you if it's working.
ONBOARDING FLOW: PRODUCT ANALYTICS FRAMEWORK KPI HIERARCHY (not all equal) PRIMARY (measure onboarding success) → D7 Return Rate (definition: any-action within 7 days of signup): TARGET 68% (vs 44% baseline old flow) → Time-to-first-action: TARGET 23 minutes (vs 38 min old flow) SECONDARY (diagnose where flow breaks) → Step completion rates: Intro=98%, Connect app=64%, Create first agent=22%, Invite team=8% → Cohort funnel: Week-1 vs week-2 signups show 12-point D7 drop (investigate seasonal) TERTIARY (longer-horizon feature adoption) → Onboard-to-create-skill: 72% of D7 returners create 1+ skill by D30 → Feature-adoption velocity curve: Skills created plateaus day 9; investigate UX barrier around agent-config step COMPARISON (old vs new) → Cohort retention analysis: old flow D30=34%, new flow D30=51% (16-point lift, p<0.001) → Per-onboarding-step confidence: Step 3 (Connect app) shows 2x higher abandon on mobile; A/B test mobile redesign ACTION SEQUENCE 1. Enable onboarding-step event tracking (5 events: flow_started, step_completed, step_skipped, step_errored, onboarding_finished) 2. Build dashboard: funnels by date, by device, by cohort 3. Weekly review: watch D7 return rate move; set alert if it drops below 62% 4. Monthly: run retention cohort analysis; flag any flow version with D30 loss vs prior month
About this skill
Most metric arguments are definition arguments, and most "retention went up" claims come from immature cohorts or a shifting user mix. This skill defines your product metrics exactly and reads cohort and adoption data the way an experienced analyst would.
What it does
- Writes a spec for every metric: unit, population, numerator, denominator, window, exclusions, owner
- Builds cohort retention tables that blank immature cells and compare cohorts at the same age
- Checks segment rates and shares before calling a trend (mix shift)
- Measures feature adoption over eligible, exposed users, with repeat use and depth
- Trims the KPI set to one outcome, a few inputs, and guardrails, each tied to a decision
What you get
Metric specs, a KPI tree, a cohort table with a plain read and confidence level, the segments that matter, caveats, and a recommendation.
Best for
Product managers, founders, and analysts setting up dashboards or checking whether a change really improved retention.
SKILL.md preview
---
name: product-analytics
description: Define product metrics precisely and run cohort, retention, and feature adoption analysis correctly - metric specs with event, actor, window, and denominator, retention curves that handle incomplete cohorts, adoption rates over the right eligible population, and a small KPI set where every metric maps to a decision. Use when the user says "define our activation metric", "what KPIs should our dashboard have?", "run a retention cohort", "is retention getting better?", "how is feature X being adopted?", "our DAU/MAU looks off", "write the metric definitions for our tracking plan", or "is this trend real or noise?".
---
… (sign up to view the full skill)More business & ops skills
View all Business & Ops skills →Product Practice Diagnostic
Trace a product-team symptom (unused features, roadmap churn, priority fights) to the broken stage of the loop and fix one thing first.
Feature Launch Go-to-Market Plan
Plan one feature launch: tier, plan access, audience messages, readiness gates for support, sales, billing, and docs, and success metrics.
Pricing Tier and Packaging Diagnosis
Finds why your pricing tiers underperform and redesigns value metric, fences, price gaps, and upgrade triggers with a migration plan.
Product Specs, Roadmaps, and Research Decisions
Turn customer evidence into testable PRDs, dependency-aware roadmaps, and stakeholder updates with clear decisions and tradeoffs.