app-analytics

Files (1)

• SKILL.md

SKILL.md (first lines)

# App Analytics

Expert in mobile app analytics/measurement. Goal: meaningful tracking, data interpretation, data-driven decisions.

# Assessment
1. Read `app-marketing-context.md`
2. Current analytics tools?
3. Top 3 performance questions?
4. What decisions need data?
5. Running paid acquisition? (attribution matters)

# Analytics Stack
Must have: App Store Connect (free, store metrics), Firebase Analytics (free, events/funnels), Crashlytics (free, stability)
Recommended: Mixpanel/Amplitude (product analytics/cohorts, free tier)
If subscriptions: RevenueCat (sub analytics/paywall testing, free tier)
If ads: Adjust/AppsFlyer (attribution, paid)

# ASC Metrics (free)
Impressions, Product Page Views, App Units (first-time DL), Conversion Rate (views->DL), Proceeds, Sessions, Active Devices, Retention (D1/D7/D28), Crash Rate
Sources: App Store Search, Browse, Web Referral, App Referral

# Metrics Framework
Acquisition: Impressions, TTR(taps/impressions=icon+title effectiveness), CVR(DL/views=product page), CPI(spend/installs), Organic%(organic/total)
Engagement: DAU, MAU, DAU/MAU(stickiness, >20% good), Sessions/User, Session Length
Retention: D1 25-40%, D7 10-20%, D30 5-10%, Monthly churn <5%(subscriptions)
Revenue: ARPU(rev/all users), ARPPU(rev/paying, 3-10x ARPU), LTV(ARPU x avg lifetime), Trial-to-Paid, MRR, Revenue Churn(lost MRR/start MRR)

# Event Tracking Plan (minimum)
```
onboarding_started, onboarding_step_completed(step_name,step_number), onboarding_completed, onboarding_skipped
[action]_started, [action]_completed, [action]_failed(error_type)
paywall_viewed(source,variant), trial_started(plan,source), purchase_completed(plan,price,source), purchase_failed(error_type), subscription_renewed, subscription_cancelled(reason)
session_started(source), feature_used(feature_name), content_viewed(type,id), share_tapped(type), notification_received(type), notification_tapped(type)
settings_changed(name,old,new), notification_permission(granted:bool)
```
Naming: snake_case, [object]_[action], specific not granular, include properties (no PII), consistent cross-platform

# Dashboards
Executive (weekly): downloads+rev+DAU+conversion+D1 retention+rating, all with trend %
Funnel (daily): Impressions->Page Views->Downloads->Activation->Purchase with conversion % between each
Cohort (monthly): retention curves by install date, acquisition source, country, plan

# Output
Analytics Audit: tools in use, events tracked count, key gaps, recommendations (tracking gaps, metrics to monitor, dashboards to create)
Tracking Plan: event name, trigger, properties, which tool
Metric Interpretation: benchmark comparison, trend analysis, specific actions

# Related
ab-test-store-listing, retention-optimization, monetization-strategy, ua-campaign

From asker26/claude-task-framework, MIT.

More from asker26/claude-task-framework