Matomo vs. Plausible: Privacy-Friendly Analytics Compared

Since Google Analytics fell out of favour in many EU compliance scenarios — keywords Schrems II, the DSK ruling and the ongoing grey area around consent banners — mid-sized companies have been looking for durable alternatives. Two solutions dominate the discussion in our customer projects: Matomo as a full-featured analytics stack with almost the entire scope of Google Analytics, and Plausible as a radically reduced tool focused on the metrics that actually matter. Both are open source, both can be self-hosted in Germany, and both can be operated in a way that requires no consent under Art. 6 GDPR.

This article compares the two candidates the way we set them up in customer projects: with Docker Compose, TLS via a reverse proxy, a clean backup strategy and an eye on running costs that still make sense after two years.

Architecture and Deployment

The most obvious difference lies in the technical foundation. Matomo is a classical LAMP stack: PHP 8.3, MariaDB or MySQL as the database, plus Nginx or Apache. Plausible is an Elixir/Phoenix application that uses ClickHouse as an analytical database and PostgreSQL for metadata.

For a typical Docker Compose deployment of Matomo 5.3 the minimum looks like this:

services:
  matomo-db:
    image: mariadb:11.4
    volumes:
      - matomo-db:/var/lib/mysql
    environment:
      MYSQL_ROOT_PASSWORD_FILE: /run/secrets/db_root
      MYSQL_DATABASE: matomo
    secrets: [db_root]

  matomo:
    image: matomo:5.3-fpm-alpine
    depends_on: [matomo-db]
    volumes:
      - matomo-data:/var/www/html
    environment:
      MATOMO_DATABASE_HOST: matomo-db

  web:
    image: nginx:1.27-alpine
    ports: ["127.0.0.1:8080:80"]
    volumes:
      - matomo-data:/var/www/html:ro
      - ./nginx.conf:/etc/nginx/nginx.conf:ro

Plausible Community Edition 3.x ships with its own docker-compose.yml that orchestrates four services: the Phoenix app, ClickHouse, PostgreSQL and optionally a mail container. For production we place a reverse proxy in front of both setups — usually HAProxy or Traefik — on the OPNsense or a dedicated ingress node, which rolls out Let’s Encrypt certificates and cleanly separates the databases from the internet.

Resource footprint at idle: Matomo runs comfortably below 1 GB RAM for small sites, Plausible sits at roughly 2 GB baseline because of ClickHouse and the BEAM VM. Both are easy to operate on a Proxmox VM with 4 vCPU and 4 GB RAM.

GDPR and consent — what works without a banner?

Both tools can be configured in a way that requires no consent banner — but with different amounts of effort.

Plausible is designed for this. The script sets no cookies whatsoever, processes no personal data in the legal sense and generates an anonymised session ID from a daily-rotated hash. It is GDPR- and TTDSG-compliant out of the box, confirmed among others in a legal opinion by the law firm Schuermann Rosenthal Dreyer that Plausible links on its own site.

Matomo needs a few deliberate settings for that. Required are:

• IP anonymisation of at least the last two bytes
• Disabling user ID capture
• No cookies (disableCookies configuration)
• No collection of user-agent details that would enable fingerprinting
• Respecting Do Not Track and the Global Privacy Control header

Once these items are set, the French data protection authority CNIL, among others, has cleared Matomo as “cookie-free”. Matomo documents the individual steps in detail at matomo.org/faq/general/faq_156/.

Feature comparison in practice

Area	Matomo 5.3	Plausible 3.x
Base metrics (visits, pageviews, referrer)	Yes	Yes
Event tracking	Yes, extensive	Yes, custom events + props
Funnel analysis	Yes (premium plugin)	Yes (goals + funnels)
Heatmaps / session recordings	Yes (premium plugin)	No
A/B testing	Yes (premium plugin)	No
E-commerce tracking	Yes, mature	Basic via events
Google Ads / Search Console import	Yes	No
Roles and permissions	Fine-grained rights system	Only owner / admin / viewer
Dashboard customisation	Extensive, multiple dashboards	One fixed dashboard
Reports by email	Yes	Yes
Raw data export	SQL, CSV, JSON API	CSV, JSON API
Log import (Nginx/Apache)	Yes	No
Server-side tag manager	Yes	No

In short: Matomo is the toolbox for marketing teams coming from Google Analytics who want to carry their analysis workflows over as closely as possible. Plausible is the tool for teams that want to answer exactly three questions — How many people showed up? Where did they come from? What did they look at? — and want a dashboard that requires no training to understand.

Cost: cloud vs. self-hosting

Both vendors offer a cloud variant and a self-hosting option. Prices as of August 2026:

Plausible Cloud: starts at 9 EUR/month for 10,000 pageviews, 19 EUR for 100,000, 69 EUR for 1 million, annual billing with a discount. All features are included in every plan.

Matomo Cloud (eu.matomo.cloud): starts at 23 EUR/month for 50,000 actions. Premium plugins such as heatmaps or Media Analytics cost extra or are bundled into higher tiers.

Self-hosting with Plausible Community and Matomo is free in terms of software; the cost sits in operation and maintenance. For a small to medium website we plan around 10-15 EUR/month for a virtualisation VM plus backup storage. In our customer projects we typically see the break-even point around 300,000 monthly pageviews, where self-hosting starts to make economic sense — below that, the cloud variant is usually the more relaxed choice.

Important note on Matomo self-hosting: premium features like heatmaps, A/B tests or the Search Engine Keywords Performance extension still require an annual plugin licence, even on-prem. For standard features though, plugins such as the official log analytics tool are open source and free of charge.

Operations, backup and migration paths

There are some practical operational details you should know before deciding.

Matomo archives raw data in MariaDB and aggregates it via a cron job (console core:archive) that should run every 15 minutes on larger installations. Without this cron, reports become slow. In our setups the database grows to roughly 200-400 GB per billion actions, depending on event diversity. A clean MariaDB backup with mariabackup captures the data incrementally — for actual retention we like to use TrueNAS snapshots with ZFS.

Plausible compresses events directly in ClickHouse; storage efficiency is significantly better, in our customer projects we typically see sizes of 20-40 GB per billion events. Backup is a bit trickier because ClickHouse and PostgreSQL must be captured consistently — a BACKUP TABLE in ClickHouse plus pg_dump for PostgreSQL, both into a volumes directory that a backup job then ships off to a separate target system.

A frequently overlooked aspect: migrating between the two systems is hard. Plausible cannot import log files, Matomo cannot import raw Plausible data. Whoever switches either loses history or has to keep the old instance running. We therefore recommend making the decision not on a purely technical basis but based on the team’s needs for the next three to five years.

Recommendation: who picks what?

After several dozen rollouts, this pattern has emerged in our projects:

Plausible fits when a small marketing team or a solo founder needs a dashboard that is understandable without a manual; when developers operate the system and don’t want to schedule PHP maintenance windows; when consent-banner freedom is a hard requirement; and when the requirement “metrics that matter” genuinely covers what the team needs.

Matomo fits when an existing marketing team is migrating from Google Analytics and needs segmentation, custom reports and attribution; when e-commerce tracking with cart analysis is required; when heatmaps and A/B tests should stay inside one tool; or when the IT department brings LAMP experience and wants to make active use of plugins.

For most mid-market B2B websites we look after, Plausible is the more relaxed choice — less operational effort, fewer GDPR discussions, less feature noise. For B2C, e-commerce or marketing-driven portals, Matomo is often the right choice despite its complexity.

DATAZONE supports the rollout

DATAZONE plans, installs and operates both analytics platforms — whether as a Docker setup on a Linux VM in your Proxmox environment, as a fully managed solution with monitoring and backup, or as a consulting sparring partner for your own IT. We know the pitfalls of GDPR-compliant configuration, the typical scaling limits of both systems and the migration paths away from Google Analytics. Get in contact with us — we start with a short current-state analysis and deliver a solid recommendation for your use case.