Self-Hosted Analytics: How to Choose Without Becoming the Tracker
Updated 2026-06-20 · 11 min read · By Ben Liu
Operator guidance for privacy-first analytics — metrics that matter, consent, and maintenance load.
What “good” looks like for a resource site
You need enough signal to improve navigation and guides — not a dossier on every visitor. Prioritize top landing pages, referrers, and device mix. Defer session replay until you have a concrete UX research question.
If you also run ads, keep analytics disclosures aligned with your cookie banner and privacy policy. Conflicting stories are a trust and compliance problem.
Shortlist criteria
Score tools on install footprint, upgrade pain, retention controls, export formats, and whether collection can fail open without breaking pages.
Browse the Analytics category with Docker and license filters, then validate release cadence upstream before importing production traffic.
Rollout advice
Ship aggregate page views first. Add events for guide CTAs later. Revisit retention quarterly and delete what you do not use.
Related deep-dive guides
- Plausible-style vs Umami vs Matomo: Analytics Trade-offs
Compare privacy posture, feature depth, and ops weight for self-hosted web analytics on a content site.
- Plausible-style vs Umami vs Matomo: Analytics Trade-offs
Compare privacy posture, feature depth, and ops weight for self-hosted web analytics on a content site.
- The Real Cost of Self-Hosting: Money, Time, and Attention
Line-item costs beyond hardware: power, disks, domains, backups, failure time, and when SaaS is cheaper in attention.