Self-Hosted Analytics: Plausible vs Umami vs Matomo — Honest Comparison

Why Self-Host Analytics?

Google Analytics 4 is a privacy nightmare. It sends your visitors' data to Google, requires cookie consent banners in the EU, and its interface is overwhelming for most teams. Self-hosted analytics give you accurate data, full privacy compliance, and zero dependency on third parties.

The Three Contenders

Umami

Umami is the minimalist choice. A single Docker container with a PostgreSQL or MySQL backend. It tracks page views, events, referrers, and basic device info — nothing more.

services:
  umami:
    image: ghcr.io/umami-software/umami:postgresql-latest
    environment:
      DATABASE_URL: postgresql://umami:password@postgres:5432/umami
      DATABASE_TYPE: postgresql
      HASH_SALT: your-random-salt-here
    ports:
      - "3000:3000"
    depends_on:
      - postgres
    restart: unless-stopped

Resource usage: ~50MB RAM, ~180MB Docker image.
Tracking script: 2KB gzipped.
Database growth: ~1MB per 10,000 pageviews.

Plausible

Plausible is privacy-focused with a beautiful dashboard. It uses ClickHouse for analytics storage, making it extremely fast for queries over large datasets.

services:
  plausible:
    image: ghcr.io/plausible/community-edition:latest
    environment:
      DATABASE_URL: postgres://plausible:password@postgres:5432/plausible
      CLICKHOUSE_DATABASE_URL: http://clickhouse:8123/plausible
      SECRET_KEY_BASE: your-64-char-secret-here
      BASE_URL: https://analytics.example.com
      DISABLE_REGISTRATION: invite_only
    depends_on:
      - postgres
      - clickhouse
    ports:
      - "8000:8000"

  clickhouse:
    image: clickhouse/clickhouse-server:latest
    volumes:
      - clickhouse_data:/var/lib/clickhouse

Resource usage: ~200MB RAM (Plausible) + ~500MB RAM (ClickHouse).
Tracking script: under 1KB gzipped.
Database growth: ~500KB per 10,000 pageviews thanks to ClickHouse compression.

Matomo

Matomo (formerly Piwik) is the feature-complete option. It matches Google Analytics feature-for-feature — funnels, cohorts, custom dimensions, A/B testing, heatmaps, session recordings.

services:
  matomo:
    image: matomo:latest
    environment:
      MATOMO_DATABASE_HOST: mariadb
      MATOMO_DATABASE_USERNAME: matomo
      MATOMO_DATABASE_PASSWORD: password
      MATOMO_DATABASE_DBNAME: matomo
    volumes:
      - matomo_data:/var/www/html
    ports:
      - "8080:80"

  mariadb:
    image: mariadb:11
    environment:
      MYSQL_ROOT_PASSWORD: root-password
      MYSQL_DATABASE: matomo
      MYSQL_USER: matomo
      MYSQL_PASSWORD: password

Resource usage: ~150MB RAM (PHP-FPM) + ~200MB (MariaDB).
Tracking script: ~22KB gzipped.
Database growth: ~5MB per 10,000 pageviews.

Feature Comparison

Custom Event Tracking

Umami

// Track a custom event
umami.track("signup-click", { plan: "pro", source: "landing-page" });

Plausible

// Track a custom event
plausible("Signup", { props: { plan: "pro", source: "landing-page" } });

Matomo

// Track a custom event
_paq.push(["trackEvent", "Signup", "Click", "Pro Plan"]);

Our Recommendation

• Choose Umami if you want simple, lightweight analytics with minimal resource usage. Best for blogs, documentation sites, and small SaaS apps.
• Choose Plausible if you want beautiful dashboards, goal tracking, and can handle the ClickHouse dependency. Best for marketing teams and content sites.
• Choose Matomo if you need full GA4 replacement with funnels, cohorts, and e-commerce tracking. Best for large applications and enterprises.

At TechSaaS, we run Umami (analytics.techsaas.cloud) because it fits our philosophy — minimal, fast, privacy-respecting. For clients who need more features, we deploy Plausible.

Need help choosing and deploying analytics? Contact [email protected].