Google Analytics is the default analytics choice for most new sites. It is free and easy to add, but it comes with two costs that tend to surface later. The first: under GDPR (the EU's General Data Protection Regulation, a privacy law requiring websites to get user permission before tracking them with cookies), many European businesses must show a consent popup asking visitors to agree to tracking. The second: your visitor data flows to Google's servers. Plausible removes both of those issues.
| Fact | Value |
|---|---|
| What it is | Privacy-first, cookieless web analytics |
| Built by | Plausible Analytics (open-source) |
| License | AGPL-3.0 (free to self-host) |
| Platforms | Docker, Plausible Cloud (EU-hosted) |
| Requires | Docker; Postgres and Clickhouse databases for self-hosting |
| Install method | Docker Compose or managed cloud subscription |
| Verdict | Use the cloud plan ($9/month) if you want zero setup; self-host if you need full data control |
What Plausible actually is
Plausible is an open-source web analytics tool with 28,175 stars on GitHub. It tracks page views, visitor counts, referrers (the websites and search engines that sent people to you), UTM parameters (tags added to URLs to track which campaign or newsletter link a visitor clicked), and goal completions (custom events you define, such as a button click or a form submission) without setting any cookies in the visitor's browser.
Most web analytics tools place cookies (small files stored in a visitor's browser that enable cross-session tracking) to identify returning visitors. Cookies trigger GDPR consent requirements in Europe. Plausible sidesteps this by counting unique visitors using a daily one-way hash (a fingerprint computed from anonymized device data that cannot be reversed to identify anyone), so no personal data is ever collected.
The tracking script is under 1 KB, versus over 70 KB for Google Analytics. That gap has a measurable effect on page load speed and search ranking.
Plausible offers two deployment paths. The managed cloud product at plausible.io starts at $9 per month for up to 10,000 monthly page views, with EU-based data storage. The self-hosted version runs under AGPL-3.0 (a copyleft open-source license: free to use, but any modifications you distribute must be shared back under the same license) and is free to run on your own infrastructure.
What you get after installing it
The dashboard is a single screen. There are no sub-menus, no "views," no configuration segments to set up before you see anything useful. You open the dashboard and your data is there.
Traffic overview shows real-time visitor counts alongside historical trends for page views, unique visitors (each person counted once per day regardless of how many pages they read), bounce rate (the percentage of visitors who leave after viewing only one page), and session duration.
Referrers and UTM tracking show where your traffic comes from. If you send a newsletter and want to know how many clicks came from it, you add a UTM tag to the link (a short label like ?utm_source=newsletter) and Plausible groups those visits automatically.
Goal tracking lets you define custom events. You add a small snippet to any button or link, and Plausible counts how many visitors trigger it. This covers conversion tracking (for example, counting how many people click a pricing page call-to-action button, or complete a signup form) without cookies.
Funnel analysis lets you define a sequence of pages or goals and see where visitors drop off. If your onboarding flow has three steps, you can track how many people complete each one.
Public dashboards let you share your analytics with anyone via a link, without giving them an account. Useful for content sites that want to show traffic openly, or for agencies reporting to clients.
Email reports send a weekly digest to your inbox so you do not need to log in to see how the week went.
Multi-site support lets one account cover as many sites as you want, up to the combined page view limit of the plan you are on.
The install experience
If you choose the cloud product, the setup is four steps: create an account at plausible.io, add your site, copy the script tag into your site's HTML head, and you are done. The script is one line. Plausible starts showing data within minutes of the first page view. There is nothing to configure before data appears.
If you choose self-hosting, the setup is more involved. Plausible requires two databases running simultaneously: Postgres (a general-purpose relational database used to store accounts, sites, and settings) and Clickhouse (a column-oriented database, meaning a type of database optimized for running fast aggregate queries over large numbers of events, which is how Plausible stores and queries page view data). Running two databases in parallel is not unusual for a production web service, but it is more infrastructure overhead than most comparable analytics tools require.
The official install method is Docker Compose (a tool that launches multiple containers from a single configuration file). The Plausible self-hosting documentation provides a ready-made docker-compose.yml that starts both databases and the application together. On a VPS (a virtual private server, a rented Linux machine you administer yourself) with Docker installed, the steps are: clone the configuration repository, edit the environment file with your domain name and a secret key, and run docker compose up.
One real constraint: Clickhouse needs memory. Plausible recommends at least 2 GB of RAM; 4 GB runs more comfortably. A $12 to $20 VPS from Hetzner or DigitalOcean is enough. If you are already running a self-hosted stack on a VPS, you can co-locate Plausible there if the RAM headroom allows it.
Where it fits and what to compare it to
Google Analytics 4 (free) is deeply integrated with Google Search Console and Google Ads and has the widest feature set of any option here. The tradeoffs: cookie-based tracking that requires a GDPR consent banner, a complex interface, and visitor data flowing to Google. For a European audience or a content site where consent banners suppress engagement, Plausible is the cleaner replacement.
Umami (github.com/umami-software/umami) is the closest open-source alternative: cookieless, self-hostable, simple dashboard, MIT licensed (a permissive open-source license with no copyleft requirements). Umami requires only a single Postgres database, making the self-hosted setup meaningfully simpler. If minimizing infrastructure overhead is a priority, evaluate Umami alongside Plausible.
Fathom Analytics (proprietary SaaS, starts at $15 per month) takes a nearly identical privacy-first approach with a cookieless method and a clean dashboard. The gap versus Plausible is the trust model: Plausible's code is auditable on GitHub; Fathom's is not. There is no self-hosting option.
Matomo (open-source, with a paid cloud option) aims to replace Google Analytics feature for feature, including heatmaps, session recordings, and A/B testing. If you need GA4-level depth and want to own the data, Matomo is the path. If you want a clean summary dashboard with minimal setup, Plausible is the better fit.
Verdict
Use the cloud plan if you want to be done in five minutes. At $9 per month for up to 10,000 monthly page views, the trade-off of paying a small subscription to avoid running your own Postgres and Clickhouse setup is worth it for most solo operators and small teams. The EU data residency and the no-cookie architecture mean you can drop it into any site without touching your consent management setup.
Self-host if you need full data control or if you are already running a self-hosted stack and want analytics that live on your own infrastructure. The two-database requirement (Postgres plus Clickhouse) is the real complexity cost here. It is not prohibitively hard for anyone who has run a Docker Compose stack before, but it is a step up from single-database tools.
The case against Plausible is narrow: if you need the deep segmentation, audience exports, and ad network integration that Google Analytics provides, Plausible is not a replacement. It is a summary dashboard, not a full behavioral analytics suite. If that limitation does not apply to your use case, the 1 KB script, the clean interface, and the no-consent-banner architecture make it a natural fit for any site where reader experience and data minimization matter. The 28,175 GitHub stars and an active cloud business both signal a project with staying power.