Swetrix is a cookieless analytics platform that started as a privacy-friendly replacement for Google Analytics and has grown into a broader product-analytics tool. The self-hosted edition is AGPL-3.0 and runs as a Docker stack; the company also sells a hosted service. For a self-hoster the interesting part is the breadth: traffic reporting, but also page performance, JavaScript errors, funnels, custom events and goals in one place.

Key capabilities

• Traffic analytics — visitors, sessions, page views, bounce rate, session duration, referrers, countries, devices and top pages, sampled at full resolution.
• Performance monitoring — real page load timings from actual visitors rather than synthetic checks, so slow pages show up next to the traffic that hit them.
• Error tracking — JavaScript errors grouped by page, browser and device.
• Custom events and goals — track signups, purchases or button clicks and set conversion goals against them.
• Funnels and session analysis — follow a multi-step flow and inspect individual sessions when a funnel step misbehaves.
• Feature flags and experiments — roll features out gradually and run A/B tests against the same data.
• Alerts and bot filtering — notifications on traffic spikes, errors or custom events, plus filters that keep datacentre traffic and spam referrers out of the numbers.

Self-hosting it

A Swetrix installation is a small set of containers — the API, the frontend and ClickHouse for the events — brought up with Docker Compose, with the tracker served from your own domain. It is cookieless, so no consent banner is needed for the analytics itself, and it does not build cross-site profiles. Some of the newer capabilities on swetrix.com, notably the CAPTCHA service and some AI-assisted analysis, are primarily part of the hosted product, so check the self-hosting documentation before you plan around them.

Alternatives in the directory

• Umami — the minimal cookieless option.
• Plausible Analytics — polished aggregate analytics without the product-analytics layer.
• PostHog — much deeper product analytics, considerably heavier to run.