Web Analytics Tools:
The 2026 shortlist (21 tested)

Why try a Google Analytics alternative?

GA4 is free, but the cost shows up elsewhere. The cookie banner kills 5–15% of conversions on landing pages. Sampling kicks in for properties under 10 million events. The interface punishes anyone who learned Universal Analytics — custom dimensions, audiences, and Explorations all behave differently. And in 2022 the French CNIL ruled that running GA on EU traffic without additional safeguards is incompatible with GDPR — Italy's Garante and Austria's DSB followed.

A privacy-first analytics tool sidesteps most of that. The seven tools above are all cookieless, store no personal data, and most claim no consent banner is needed in the EU. Their dashboards are an order of magnitude simpler — pageviews, referrers, top pages, devices, and country — and you can install them in under five minutes with a single script tag.

They aren't drop-in replacements for product analytics (PostHog, Mixpanel) or full marketing attribution stacks (GA4 + Google Ads). But for content sites, indie SaaS, agencies, and EU-based bootstrappers, they cover the 80% of analytics that actually drives decisions.

Cookieless analytics: how does it work without cookies?

Traditional analytics drops a third-party cookie to tell visits apart. Cookieless tracking uses a server-side hash instead — a one-way function of IP + User-Agent + a daily rotating salt. The hash gets discarded when the salt rotates (typically every 24 hours), so no individual is tracked across days.

All seven tools we reviewed are cookieless by default:

• Plausible — daily salt rotation
• Fathom — daily salt rotation, plus EU Isolation routing
• Pirsch — 24-hour rotation, German jurisdiction
• GoatCounter — 8-hour rotation (most aggressive in the field)
• Umami — monthly default rotation (configurable)
• Simple Analytics — IPs dropped entirely, referrer-based unique detection
• Matomo — opt-in cookieless mode (cookies are on by default)

The trade-off: cookieless tracking can't reliably tell a returning visitor from a new one across days. For most marketing dashboards that's fine — you care about traffic source and conversion rate, not individual user journeys.

Self-hosted analytics: which tools can you run on your own server?

Self-hosted analytics means you run the tool on your own infrastructure — VPS, Docker container, or Kubernetes cluster. Your visitor data never reaches a third-party SaaS. No DPA, no Schrems-II concern, no per-pageview pricing. The trade-off is operations: backups, upgrades, TLS, scaling.

Of our seven tools, four are genuinely self-hostable:

• Matomo — full GPL-3.0 codebase, PHP + MySQL stack, official Docker image
• Umami — MIT license, Next.js + PostgreSQL, single Docker Compose file
• Plausible — AGPL-3.0 Community Edition, Elixir + ClickHouse
• GoatCounter — EUPL-1.2, single Go binary on SQLite (lightest stack)

Pirsch publishes its core as an AGPL Go library but the dashboard is closed-source — full self-host needs custom work. Fathom and Simple Analytics are SaaS-only.

For most operators we recommend self-hosting if you have ops capacity and want zero recurring cost, or EU-hosted Cloud if you'd rather pay $9–29/mo and never think about updates.

Popular head-to-head comparisons

The most-searched pairs in our directory — opinionated side-by-side breakdowns with full feature matrices and editorial verdicts: