Umami - Privacy-Focused Self-Hosted Web Analytics

Umami is a privacy-focused web analytics platform positioned as a simpler alternative to Google Analytics. The project is open source, maintained on GitHub, and its own documentation describes it as cookie-free, without cross-site tracking or collection of personal data.

For teams that want traffic analytics without moving visitor data into a large advertising ecosystem, Umami sits in a practical middle ground: lightweight enough for small sites, but with enough analysis features to be useful beyond a basic pageview counter.

Why privacy analytics keeps coming up

Many site owners still need to know which pages work, where visitors arrive from, and whether campaigns or product pages are performing. The tension is that conventional analytics stacks can be too heavy, too invasive, or too opaque for sites that mainly need operational insight.

Umamiâs answer is a focused analytics product that keeps the user experience close to the essentials. Its docs emphasize no cookies, no fingerprinting, no tracking across sites, and the option to self-host so the analytics data stays on your own infrastructure.

What Umami provides

The repository describes Umami as a simple, fast, privacy-focused alternative to Google Analytics. The official documentation expands that into a broader analytics feature set, including core website metrics, custom events, and more advanced views.

In practical terms, the source material points to:

pageviews, visitors, bounce rate, session duration, referrers, browsers, operating systems, devices, and countries;
custom events for interactions such as button clicks or form submissions;
funnels, journeys, retention analysis, goals, UTM campaign tracking, and cohort breakdowns;
teams, shared website access, roles, and a REST API.

That makes Umami more than a minimal hit counter, but still less sprawling than a full product analytics suite designed around every possible growth, attribution, and experimentation workflow.

Self-hosting and deployment shape

Umami is comparatively straightforward to place on your own server. The GitHub README lists Node.js 18.18 or newer and PostgreSQL 12.14 or newer for source installs, and it also provides Docker images and a Docker Compose path that runs Umami with PostgreSQL.

For a small self-hosted setup, the operational model is easy to understand: a web application, a PostgreSQL database, environment configuration, and a reverse proxy or direct port exposure depending on how you publish it. That does not remove the normal responsibilities of self-hosting â backups, updates, TLS, monitoring, and database maintenance â but the stack is conventional and familiar to many web developers.

Umami vs Plausible for self-hosters

Plausible is the most obvious comparison because it is also privacy-focused, open source, and available in a self-hosted Community Edition. The trade-off is not simply âbetterâ or âworseâ; it depends on how much operational complexity you want to accept.

For ease of use, both products aim for clean dashboards and simple analytics. Umamiâs self-hosting path is often easier to reason about because the basic stack centers on the app plus PostgreSQL. Plausible Community Editionâs official self-hosting repository uses Docker Compose and includes ClickHouse; its prerequisites call out Docker, Docker Compose, CPU instruction support required by ClickHouse, and at least 2 GB of RAM to avoid out-of-memory issues. That makes Plausible CE feel a little more infrastructure-heavy for the smallest VPS deployments.

For deployment, Umamiâs Docker Compose route is direct, and its source install requirements are clearly stated in the README. Plausible CE also has an official Compose setup, but the projectâs own self-hosting page stresses that self-hosters are responsible for installation, maintenance, upgrades, server capacity, uptime, backups, security, stability, and loading time. Plausible also distinguishes between its continuously updated hosted cloud product and the Community Edition release line.

For popularity, GitHub gives one visible but imperfect signal. The Umami repository currently shows a much larger star count than the Plausible Community Edition repository, though that comparison is not perfectly apples to apples because Plausibleâs hosted product and CE packaging are presented separately. The safer conclusion is that both are well-known in the privacy analytics space, while Umamiâs main repository currently has stronger GitHub visibility.

For resources, Umami looks attractive for low-maintenance self-hosting because PostgreSQL is often already part of a small web stack. Plausibleâs ClickHouse-based setup can be powerful for analytics workloads, but it raises the baseline resource and operations profile. On a very small server, that distinction matters.

Best-fit scenarios

Umami fits site owners, agencies, publishers, and small product teams that want a privacy-conscious analytics layer with a manageable self-hosting footprint. It is especially appealing when you already run PostgreSQL-backed services and want analytics data under your own control.

It is also a sensible option for developers who want a dashboard that is easier to explain to non-technical colleagues than a sprawling enterprise analytics tool. The feature set is broad enough for campaign tracking and event analysis, but the product identity remains focused on understandable web analytics.

Caveats before adopting it

Self-hosting analytics is still self-hosting. You need to keep the service patched, watch disk growth, back up the database, secure the admin interface, and make sure the tracking endpoint remains reachable without becoming an attack surface.

There is also a strategic caveat: if you need enterprise attribution modeling, very deep ecommerce reporting, guaranteed managed uptime, or hands-off compliance operations, a hosted analytics service may be a better operational fit. Umami can reduce dependency on third-party analytics platforms, but it does not remove the need to operate software responsibly.

Editorial verdict

Umami is one of the more practical choices for self-hosted, privacy-focused web analytics. Its appeal is not that it promises every metric imaginable, but that it keeps the core analytics problem understandable: collect useful website statistics, avoid invasive tracking patterns, and let operators keep control of their data.

Compared with Plausible Community Edition, Umami looks especially compelling when ease of self-hosted deployment and modest resource needs are high priorities. Plausible remains a strong privacy analytics brand with a polished product direction, but the self-hosted CE path appears more demanding because of its ClickHouse-based stack and the maintenance responsibilities called out in its own documentation.