Skip to content

admin-cohort-analysis.md

Latest commit

 

History

History
50 lines (35 loc) · 1.85 KB

admin-cohort-analysis.md

File metadata and controls

50 lines (35 loc) · 1.85 KB
Raw

Admin Cohort Retention Analysis

A triangular cohort retention table showing what percentage of users who signed up in a given month were still active (logged in) in subsequent months. A critical SaaS growth metric for identifying activation and retention drop-offs.

Accessing the Dashboard

URL: /admin/cohort-analysis (requires is_superadmin)

How It Works

  • The last 6 signup cohorts (calendar months) are displayed.
  • For each cohort, 4 retention columns are computed:
    • Month 0: always 100% (the cohort baseline).
    • Month 1, 2, 3: percentage of cohort members who made at least one successful login in that calendar month.
  • Future months are shown as — and excluded from averages.
  • An average row at the bottom summarises retention across all cohorts.

Colour Coding

Colour Threshold
Green ≥ 70%
Amber 40–69%
Red < 40%

Architecture

File Role
app/Http/Controllers/Admin/CohortAnalysisController.php Data aggregation + Inertia render
resources/js/pages/admin/cohort-analysis.tsx Triangular cohort table UI
resources/js/layouts/admin-layout.tsx Nav entry (Cohort Analysis)
routes/web.php GET /admin/cohort-analysis route

Data Sources

  • Cohort membership: users.created_at grouped by calendar month.
  • Retention activity: login_activities table, filtered by is_successful = true and the target month window, using DISTINCT user_id.

Testing

Test file: tests/Feature/Admin/CohortAnalysisTest.php

  • Superadmin can access the page
  • Non-admin is forbidden
  • Response has correct Inertia component + cohorts prop shape
  • Month 0 is always 100% in every cohort
  • Actual login activity is reflected in retention percentages

Run: php artisan test --compact tests/Feature/Admin/CohortAnalysisTest.php