Rybbit

Open-source privacy analytics, honestly reviewed. No cookie banners, no data brokers, no marketing fluff.

TL;DR

• What it is: Open-source (AGPL-3.0) web analytics platform — think Google Analytics with zero cookies, full self-hosting, and a UI that doesn’t require a PhD to read [1][3].
• Who it’s for: Non-technical founders and small teams tired of GA4’s complexity, Plausible’s missing features, and the general anxiety of handing visitor data to Google [2][3].
• Cost savings: Plausible starts around $9/mo and climbs with pageview volume. Google Analytics is free but your data goes to Google. Rybbit self-hosted runs on a $5–10/mo VPS with no per-pageview pricing and session replays included at the software level.
• Key strength: The only tool in the privacy-analytics category that combines a clean beginner UI with session replays, user journeys, retention analysis, funnels, and error tracking — all self-hostable and all open-source [3].
• Key weakness: AGPL-3.0 license means you can’t embed it in a commercial product without open-sourcing your own code. Session replays are gated behind the Pro tier on the cloud version. Newer project — only 11,844 GitHub stars versus Plausible’s years of community trust.

---

What Is Rybbit

Rybbit is a web and product analytics platform built as a privacy-first replacement for Google Analytics. You drop one line of JavaScript into your site, and it starts recording sessions, pageviews, user flows, conversions, and web performance — with no cookies, no user fingerprinting beyond session anonymization, and no data leaving your server if you self-host.

The project describes itself in the GitHub README as “an open-source and privacy-friendly alternative to Google Analytics that is 10x more intuitive,” which is a bolder claim than most analytics tools make. Whether that’s true depends on your baseline: if your baseline is GA4, the bar is remarkably low. If your baseline is Plausible (which is already clean), it’s a fairer fight [README][3].

Technically, Rybbit runs on ClickHouse for event storage and PostgreSQL for application data — the same stack behind most high-performance analytics platforms [2]. That architecture choice matters: ClickHouse handles analytics queries at scale in ways that general-purpose databases don’t. It’s why Rybbit can offer real-time dashboards and complex funnel queries without the query timeouts that plague simpler implementations.

The project is hosted on EU infrastructure (Germany), which matters for GDPR compliance — not just as a marketing claim but as a practical fact about where data physically lives [homepage]. As of this review, it has 11,844 GitHub stars, is trusted by “4,000+ organizations,” and hit Product Hunt’s daily top post.

---

Why People Choose It

The migration stories in the third-party reviews follow a consistent pattern: GA4 → [intermediate tool] → Rybbit. The intermediate tool is usually Plausible or Umami.

From GA4: The complaints are predictable but genuine. GA4’s interface was redesigned for Google’s ad business, not for site owners trying to understand basic traffic. One reviewer [3] lists the specific frustrations that pushed him out: a confusing UI that requires extensive configuration to get useful data, aggressive data sampling on free plans, bot traffic getting recorded as real sessions, and poor data capture from visitors running ad blockers. GA4 is also not privacy-friendly by any stretch — every event goes to Google’s servers and feeds their advertising machine. For founders who’ve been burned by GDPR compliance reviews or just don’t want to explain to EU customers why their browsing behavior is in Mountain View, GA4 is increasingly a liability [1].

From Plausible: This is where the Rybbit pitch gets more interesting. Plausible is already privacy-first, already cookieless, already simple. The switch from Plausible to Rybbit is about feature gaps. Akash Rajpurohit [2] switched from Umami (citing slow development and feature limitations), but the Plausible comparison is similar: no free tier on Plausible, no session replay, no retention analysis, no error tracking, and the self-hosted community edition offers materially fewer features than the paid cloud version [3]. Rybbit’s self-hosted version is the same codebase as the cloud version — you don’t get a stripped-down open-core where the good parts are paywalled.

The travis.media review [3] summarizes the competitive position cleanly: Rybbit provides everything Plausible does plus session replays, user journeys, and retention analysis, at a lower price point for higher traffic volumes. That’s a strong value proposition for any founder who has bounced off Plausible’s pricing page and wondered why session replay requires a separate $X/month Hotjar subscription.

---

Features

Based on the README, website, and independent reviews:

Core analytics:

• Sessions, unique visitors, pageviews, bounce rate, session duration [README]
• Real-time dashboard with live visitor map [1]
• 3-level geographic tracking (country → region → city) with 3D globe visualizations [README]
• Device, browser, OS, and referrer breakdowns [1]
• Advanced filtering across 15+ dimensions [README]

Behavioral analytics:

• Session replays — watch individual user recordings to see exactly where they clicked, scrolled, and dropped off [2][3]
• User journey mapping — see the actual navigation paths users take from landing page to conversion [README][3]
• Funnel analysis with configurable steps [README][3]
• Retention analysis (cohort-style) [README]
• Custom events with JSON properties — track sign-ups, purchases, or any interaction with structured metadata [README]

Performance and errors:

• Web Vitals monitoring (LCP, CLS, INP, FCP) [1][README]
• Error tracking [3][README]

Operations:

• Public dashboards — share a read-only analytics view with clients or the public [README]
• Organizations and multi-site support [README]
• Goal tracking [README]
• Bot filtering, automatic [homepage]
• API access [homepage]
• Data export [homepage]
• Email reports [homepage]

What you won’t find:

• Direct data import from other platforms — if you’re migrating from Plausible or GA4, your historical data stays behind [2]
• SSO and single sign-on, except in the Enterprise tier [homepage pricing]
• White-labeling, except at Enterprise [homepage pricing]

---

Pricing: SaaS vs Self-Hosted Math

Rybbit Cloud:

• Free tier: 3,000 monthly pageviews, limited to trial evaluation [1][3]
• Standard: $13/month — up to 5 websites, up to 3 team members, custom events, funnels, goals, journeys, web vitals [homepage pricing]
• Pro: $26/month — unlimited websites, unlimited team members, session replays, 5-year data retention, priority support [homepage pricing]
• Enterprise: Custom — SSO, infinite data retention, dedicated isolated instance, on-premise installation, custom features, white-labeling [homepage pricing]

Note: pricing above is for the 100K monthly pageviews tier. The slider on the pricing page goes up to 50M+ pageviews, so costs scale. Exact higher-tier pricing not listed publicly.

Self-hosted (AGPL-3.0):

• Software: $0
• VPS: $5–10/month (Hetzner, Contabo, or DigitalOcean)
• Your time to set it up

Plausible for comparison:

• No free tier (trial only)
• Starts around $9/month for 10K pageviews and climbs with traffic volume
• Self-hosted Community Edition exists but with reduced features [3]

Google Analytics:

• Free, but your visitors’ behavioral data goes to Google
• GA4 360 (enterprise) runs into thousands per month

Concrete math for a typical small SaaS founder:

Say you’re running 500K pageviews/month across 3 websites. On Plausible that lands somewhere around $39–69/month depending on the plan. On Rybbit Standard, you’d need to check the slider pricing, but the $13/month entry point with 5 websites is notably aggressive at the low end. Self-hosted on a $6 Hetzner VPS: $6/month for unlimited pageviews across unlimited sites. The session replay feature alone — which would cost you a separate Hotjar plan if you’re on Plausible — is included in the Pro tier or in the self-hosted codebase.

One important caveat: Rybbit is newer and the pricing may not have stabilized. The current public numbers are attractive, but early-stage SaaS pricing often changes as companies find their margin ceiling.

---

Deployment Reality Check

Self-hosting Rybbit is genuinely practical for someone who’s run a Docker container before. The stack is Docker Compose with ClickHouse and PostgreSQL bundled — you’re not wiring up six services manually [2][README].

What you need:

• A Linux VPS with at least 2GB RAM (ClickHouse has a heavier footprint than Plausible or Umami)
• Docker and docker-compose
• A domain and reverse proxy (Caddy or nginx) for HTTPS
• About 30 minutes if you’ve done this before [3]

What Akash Rajpurohit does in practice [2]: He recommends a hybrid setup — keep the Rybbit dashboard internal (accessible only via Tailscale) while exposing just the API tracking endpoint through Cloudflare Tunnels. This means the analytics collection endpoint is publicly reachable (so your tracking script can phone home), but the full dashboard requires VPN access. Smart paranoia for anyone running a homelab.

What can go sideways:

• No direct data import from GA4, Plausible, or Umami. You start fresh — your historical data doesn’t move [2]. For an existing site with years of analytics history, this is a real cost.
• ClickHouse is heavier than most analytics tools’ storage engines. On a 1GB RAM VPS you’ll likely hit memory pressure. Budget at least 2GB, preferably 4GB [architecture context].
• AGPL-3.0 license has implications if you’re building something on top of Rybbit. If you want to embed analytics in a SaaS product and not open-source your code, AGPL blocks that. This is meaningfully different from Plausible’s MIT license for the tracking script, or how n8n’s “Fair-code” license works. If you’re just tracking your own site, it doesn’t matter at all.

Realistic time estimate: 30–60 minutes for a technical user. 2–3 hours for a non-technical founder following a guide, including domain setup. If you’ve never touched a Linux server, either learn Docker Compose basics first or pay someone to deploy it once.

---

Pros and Cons

Pros

• Full feature parity on self-hosted. The self-hosted version is the same codebase as the cloud version — no features gated behind “upgrade to cloud” [3]. This is rare in analytics tools.
• Session replays included without a separate tool. On the cloud, it’s Pro-tier. On self-hosted, it’s part of the package. That alone eliminates a Hotjar or FullStory subscription [2][3].
• GA4-level behavioral features with Plausible-level simplicity. Funnels, user journeys, retention analysis, and error tracking in one dashboard — features that GA4 has but buries behind five layers of configuration [3].
• Genuinely cookieless. No consent banners, automatic GDPR/CCPA compliance by architecture, not by policy [1][homepage].
• EU infrastructure. Hosted in Germany, which matters for European businesses dealing with data residency requirements [homepage].
• Generous free tier. 3,000 monthly events for evaluation, with no credit card required [1][3].
• Clean, fast UI. Reviewers consistently mention the interface as the primary reason they stayed [2][3]. ClickHouse makes queries fast.
• Open source with self-hosting parity. No open-core gating of key features [README].

Cons

• AGPL-3.0, not MIT. You can self-host for your own site freely. You cannot embed Rybbit in a commercial product or SaaS without open-sourcing your code. This eliminates a “resell analytics as a feature” use case that more permissively licensed tools allow.
• Session replays are Pro-only on cloud. At $26/month, Pro isn’t expensive, but it’s not the $13 Standard you might have budgeted. Self-hosters get it free [homepage pricing].
• No historical data migration. If you’re switching from Plausible or GA4, your historical analytics don’t transfer. You start over [2].
• Newer, smaller community. 11,844 GitHub stars is solid for a young project, but it’s a fraction of Plausible’s user base or GA4’s documentation depth. You’ll find fewer “how do I do X with Rybbit” answers on Stack Overflow.
• ClickHouse has a heavier memory footprint than alternatives like SQLite-backed Umami. Budget more RAM for the VPS than you’d expect [2][architecture].
• SSO only at Enterprise. If you’re running Rybbit for a team and need centralized identity management, you’re paying custom Enterprise pricing or managing users manually [homepage pricing].
• No vendor-backed migration tooling. The setup is well-documented, but if you want smooth migration from a specific tool, you’re on your own [2].

---

Who Should Use This / Who Shouldn’t

Use Rybbit if:

• You’re paying for Google Analytics 360, Hotjar, and a separate funnel tool and want to consolidate into one self-hosted stack.
• You’re a non-technical founder who’s bounced off GA4’s configuration complexity and wants session replays without a separate subscription.
• You need GDPR/CCPA compliance without cookie banners — and you need to prove it to an auditor, not just assert it.
• You’re comfortable with basic Docker deployment (or willing to learn, or pay someone once to set it up).
• You have EU customers who ask what happens to their browsing data.
• You’re currently on Plausible’s paid plan and feel like you’re paying for half the features you need.

Skip it (stay on Plausible) if:

• Your analytics needs are genuinely simple — pageviews, bounce rate, top referrers. Plausible does this better than anyone, is battle-tested, and has years of community trust. Rybbit is solving a different problem.
• You value a mature, stable product over a feature-rich newer one. Plausible has been around longer and has more production deployments.
• You need MIT-licensed code to embed in your own product.

Skip it (use PostHog) if:

• You’re building a product analytics setup with event tracking, A/B testing, feature flags, and user profiles tied to authenticated users. PostHog is the right tool for that problem. Rybbit is primarily website analytics with some product analytics overlap.

Stay on GA4 if:

• Your compliance team requires the certification and indemnification that comes with Google Workspace. GA4 has legal paperwork that self-hosted tools don’t.
• You need the integration with Google Ads, Search Console, and Looker Studio. That ecosystem is Google’s moat and nothing open-source touches it.

---

Alternatives Worth Considering

• Plausible — the cleaner, simpler, more mature privacy-analytics option. No session replays, no retention analysis. Starts around $9/month; self-hosted community edition has fewer features. Good choice if simplicity is the priority over feature depth.
• Umami — open-source (MIT), simpler feature set, slower development cycle. The tool Akash Rajpurohit was running before Rybbit [2]. Good for pure pageview analytics on a tight budget.
• Fathom — privacy-first, EU-based, closed source SaaS. Clean and fast, no self-hosting option.
• PostHog — open-source product analytics with session replays, feature flags, A/B testing, and heatmaps. More powerful for product instrumentation, heavier to self-host, more complex UI. The right choice if you’re tracking authenticated users and product-level events, not marketing site traffic.
• Matomo — the oldest self-hosted Google Analytics alternative. Extremely feature-complete, can be configured to be privacy-compliant, but the UI is dated and the self-hosted setup is heavier than Rybbit’s. Worth considering if you need the most comprehensive data model.
• Cloudflare Web Analytics — free, cookieless, no self-hosting required. Almost no features beyond pageviews and performance. Good as a supplementary signal, not a primary analytics tool.

For a non-technical founder who needs to escape GA4 and wants session replays without five separate tools, the realistic shortlist is Rybbit vs PostHog. Pick Rybbit if you run a marketing/content site and want clean traffic analytics with behavioral context. Pick PostHog if you’re instrumenting a SaaS product and need user-level event tracking tied to authenticated accounts.

---

Bottom Line

Rybbit fills a specific gap that nobody else fills cleanly: privacy-first web analytics with session replays, user journeys, retention analysis, and error tracking, all in one self-hostable package, all open-source. Plausible is more mature but feature-limited. PostHog is more powerful but built for product instrumentation, not marketing analytics. GA4 is free but you’re the product. Rybbit sits in the middle of that map and does it without cookie consent forms or per-pageview anxiety.

The trade-offs are real: AGPL-3.0 limits commercial embedding, the project is younger than its competitors, and ClickHouse means your VPS needs real RAM. But for a founder paying $40–80/month across GA4, Hotjar, and a funnel tool, the self-hosted Rybbit math is hard to argue with. A $10 VPS, an afternoon of setup, and that entire stack consolidates into one dashboard that you own.

If the afternoon of setup is the blocker, that’s exactly what unsubbed.co’s parent studio upready.dev deploys for clients. One-time fee, done, you own the infrastructure.

---

Sources

1. Tom Martin — “Rybbit Analytics Review: Cookie-Free Google Analytics Alternative” (November 3, 2025). https://analytics-alternatives.com/rybbit-analytics-a-cookie-free-alternative-to-google-analytics/

2. Akash Rajpurohit — “Rybbit — Privacy-focused open-source analytics that actually makes sense” (September 11, 2025). https://akashrajpurohit.com/blog/rybbit-privacy-focused-open-source-analytics-platform/

3. travis.media — “The Best Google Analytics Alternative in 2026 (I Tried Them All)”. https://travis.media/blog/rybbit-best-google-analytics-alternative/

Primary sources:

• GitHub repository and README: https://github.com/rybbit-io/rybbit (11,844 stars, AGPL-3.0)
• Official website: https://rybbit.com
• Pricing page: https://rybbit.com/pricing
• Documentation: https://rybbit.com/docs
• Live demo: https://demo.rybbit.com/1