Should I self-host Plausible Analytics?

What you lose

• Their infrastructure absorbing traffic spikes — a front-page day is now your ClickHouse's problem
• Funding the developers whose code you're running (their pricing page will tell you the same, correctly)
• Managed email reports and uptime — small, but it's on your pager now

What you're paying Plausible Analytics for

• privacy-friendly web analytics
• goals & custom events
• no pageview caps
• email/Slack reports

Ranked alternatives

1. Plausible CE

The same privacy-first analytics dashboard the vendor sells, on your own domain, with no pageview meter counting against you.

plausible/analytics · ★ 28,463 · last commit 0d ago · AGPL-3.0

2. Umami

Pageviews, referrers and custom events on a hostname you own, from one container and a PostgreSQL, with no monthly event meter.

umami-software/umami · ★ 38,146 · last commit 0d ago · MIT

Hardware

Runs comfortably on a real VPS (2 GB+).

The tested compose file (our evidence, CI-booted)

# Plausible Community Edition — the rare case where the SaaS and the self-host are the
# same software. Secrets are throwaway test values: openssl rand -base64 48
services:
  plausible:
    image: ghcr.io/plausible/community-edition:v3
    container_name: plausible
    restart: unless-stopped
    command: sh -c "/entrypoint.sh db createdb && /entrypoint.sh db migrate && /entrypoint.sh run"
    depends_on:
      plausible-db:
        condition: service_healthy
      plausible-events-db:
        condition: service_healthy
    environment:
      BASE_URL: http://localhost:8000
      SECRET_KEY_BASE: "GLVzDZW04FzuS1gMcmBRVhwgd4Gu9YmSl/k/TqfTUXti7FLBd7aflXeQDdwCj6Cz"
      DATABASE_URL: postgres://plausible:plausible@plausible-db:5432/plausible_db
      CLICKHOUSE_DATABASE_URL: http://plausible-events-db:8123/plausible_events_db
      DISABLE_REGISTRATION: "false"
    ports:
      - "8000:8000"
    healthcheck:
      test: ["CMD-SHELL", "wget -qO- http://localhost:8000/api/health || exit 1"]
      interval: 5s
      timeout: 5s
      retries: 36
      start_period: 30s

  plausible-db:
    image: postgres:16-alpine
    restart: unless-stopped
    environment:
      POSTGRES_USER: plausible
      POSTGRES_PASSWORD: plausible
      POSTGRES_DB: plausible_db
    volumes:
      - plausible-pg:/var/lib/postgresql/data
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U plausible"]
      interval: 5s
      timeout: 5s
      retries: 12

  plausible-events-db:
    image: clickhouse/clickhouse-server:24.12-alpine
    restart: unless-stopped
    volumes:
      - plausible-ch:/var/lib/clickhouse
    ulimits:
      nofile:
        soft: 262144
        hard: 262144
    healthcheck:
      test: ["CMD-SHELL", "wget -qO- http://localhost:8123/ping || exit 1"]
      interval: 5s
      timeout: 5s
      retries: 24
      start_period: 20s

volumes:
  plausible-pg:
  plausible-ch:

Want a guided install instead? caniselfhostit.com/plausible has AI-agent prompts that assume a bare machine — that's their half of the stool, and it's good.

Verification

# Timed setup log: Plausible (cloud) → Plausible Community Edition

**Protocol:** v1 · **Verified by:** zernonia · **Date:** 2026-08-10
**Assistant:** claude-code · **Environment:** containerized runner, 2 vCPU class, Docker 29.3 / Compose v5.1

## Timeline

| Step | Time |
|---|---|
| Read CE hosting docs, draft compose (app + postgres + clickhouse, secret, createdb/migrate entry) | 25 min |
| Boot attempt in our verification environment: **blocked — ghcr.io unreachable** (see What broke) | 5 min |
| Compose boot + health verified by the CI runner instead (`compose-check` workflow) | — |
| Workflow review against CE docs: registration, site creation, script snippet | 10 min |
| **Total: ~40 min** | |

## Measurements

- Verification environment could not pull ghcr.io images; **boot-to-healthy for this pair is
  certified by CI**, which runs the same `scripts/check-compose.sh` gate on every change and
  weekly on schedule. Treat the CI badge, not this log, as the boot evidence.
- 3 containers; ClickHouse is the heavy one — plan ~1 GB RAM for it alone. This drives the
  `vps` hardware tier and the honest `vps_share` in the math.

## What broke

1. **CE images live on ghcr.io only.** Networks that mirror just Docker Hub (some CI setups,
   corporate proxies — and our own verification box today) can't pull them. Nothing wrong with
   the software; worth knowing before you assume any registry works from your box.
2. **First boot must run `db createdb && db migrate` before `run`** — the compose's `command`
   handles it; a bare `run` on a fresh volume exits with a confusing DB error.

## Verdict-relevant notes

- Same software as the SaaS: capability is 100% by construction, minus their managed spikes.
- The YES survives the math but barely — ClickHouse's RAM appetite makes this the thinnest
  YES margin on the site. Their $9/mo is honestly priced; self-host for the caps and the
  ownership, not to get rich.

Verdict history

• 2026-08-10: unscored → YES — Initial verdict: same software, no pageview cap; margin thinner than it looks (ClickHouse eats RAM).