Prisme Analytics

Open-source, cookie-less web analytics, honestly reviewed. No marketing fluff, just what you get when you self-host it.

TL;DR

• What it is: Open-source (AGPL-3.0) web analytics — privacy-first, cookie-less, built on Grafana for dashboards. Targets founders tired of sending visitor data to Google [3].
• Who it’s for: Small teams and solo founders who need GDPR compliance out of the box, want to own their data completely, and don’t mind a Docker-based self-hosted setup.
• Cost savings: Plausible starts at $9/mo for 10K pageviews and scales steeply. Fathom starts at $15/mo. Prisme Cloud charges €8.99/mo with unlimited pageviews and unlimited websites — self-hosted brings that to the cost of a VPS [pricing page].
• Key strength: One flat price covering unlimited everything, Grafana-backed custom dashboards, a 2kB tracking script that won’t hurt your SEO, and no cookie banner required [README][3].
• Key weakness: 123 GitHub stars at time of writing — this is early-stage software with a thin review corpus, an AGPL-3.0 license that blocks commercial embedding, and a Grafana dependency that makes self-hosting heavier than it looks [1][README].

---

What is Prisme Analytics

Prisme Analytics is a self-hosted, cookie-less web analytics service written in Go. You drop a 2kB JavaScript snippet on your site, and within minutes you get pageviews, sessions, referrer sources, UTM campaign data, and bot-filtered traffic in a Grafana dashboard. The GitHub description is honest about what it is: “High-performance, self-hosted and privacy-focused web analytics service. Written in Go” [README].

The product positions itself squarely as a Google Analytics replacement for privacy-conscious teams [3]. It’s AGPL-3.0 licensed, meaning the full source code is open, you can self-host it, modify it, and distribute it — but if you embed it in a commercial product or SaaS, you must open-source that product too. That’s the trade-off versus MIT-licensed tools.

What makes the architecture interesting is the Grafana dependency. Rather than building a custom dashboard UI from scratch, Prisme leans on Grafana for all visualization: panels, filters, custom dashboards, user management, team permissions, and multi-organization support all come from Grafana [README]. This is a smart shortcut — it gives you a battle-tested, extensible BI layer for free — but it adds operational weight when self-hosting.

The project is built by Alexandre Negrel, a solo developer, and added to AlternativeTo in August 2024 [1]. As of this review it sits at approximately 123 GitHub stars with 2 forks — by comparison, Plausible has 20K+ stars and Matomo has 20K+. That’s not a disqualifier, but it’s an important signal about community maturity and long-term maintenance risk.

---

Why people choose it

Independent third-party reviews are thin. The AlternativeTo listing shows 11 likes and 85 listed alternatives, with no user reviews or comments posted [1]. What’s available comes from the product’s own documentation, the AlternativeTo community activity (users adding it as an alternative to tools like GoodMetrics, HitKeep, and Nanolytica) [1], and the GitHub README.

The core reasons people land on Prisme instead of competing tools:

Versus Google Analytics. This is the primary comparison the product makes [3]. Google Analytics is closed-source, collects data on Google’s servers, requires consent banners in most jurisdictions, and takes 24–48 hours to process data [3][pricing page]. Prisme is open-source, processes data on your server, requires no cookie banner (cookie-less by design), and shows data in real time [README][3]. For a founder who’s uncomfortable with “my customer’s behavior is Google’s data,” this is a clean argument.

Versus Plausible and Fathom. Both are well-established privacy-first analytics services with larger communities. The Prisme differentiation is pricing structure: Plausible and Fathom charge per pageview tier, which means costs climb as traffic grows. Prisme’s cloud plan is a flat €8.99/mo for unlimited pageviews, unlimited websites, and unlimited dashboards [pricing page]. For high-traffic sites, the math flips in Prisme’s favor. For sites under 10K pageviews/month, Plausible’s $9/mo entry tier is competitive.

Versus Matomo. Matomo is the most mature open-source analytics platform, GPL-3.0 licensed, with years of production use and a large ecosystem. Prisme is simpler and lighter — the 2kB tracking script versus Matomo’s heavier footprint, and a simpler dashboard model. Matomo’s self-hosted version is free but configuration-heavy. Prisme self-hosted is lighter to reason about, though the Grafana dependency adds its own complexity [README].

On the Grafana-as-dashboard angle. The README explicitly calls out Grafana integration as a feature: user management, team management, permissions, multi-organization support, and custom dashboards all come from Grafana [README]. For teams that already run Grafana internally, this is a meaningful advantage. For teams new to Grafana, it’s an additional thing to learn.

---

Features

Based on the README and website:

Core analytics:

• Pageviews, sessions, unique visitors, bounce rate [README][website]
• Real-time data processing — no 24–48 hour lag [3][website]
• Referral sources and UTM campaign tracking [README][website]
• Bot, scraper, and spam traffic filtering — automatic [README][website]
• SPA (single-page app) support via pushState router detection [README]
• Cookie-less tracking — no consent banner required [README][3]

Tracking script:

• ~2kB JavaScript snippet (~22x smaller than Google Analytics) [README]
• Noscript fallback: a 35-byte transparent pixel GIF for Javascript-disabled visitors [README][website]

Custom analytics:

• Custom events with arbitrary data payloads for tracking business-specific actions [website]
• Custom dashboards via Grafana — funnels, histograms, graphs, geomaps [website]
• SQL query API for direct data access — no intermediary, no aggregation black box [website]
• Filtering down to single data points [website]

Infrastructure:

• Docker deployment [README][features canonical]
• REST API [features canonical]
• Grafana for all dashboard/user/team management [README]
• Performance tested at 50,000+ req/s on a consumer-grade AMD Ryzen 7 [README]
• Forever data retention — no auto-deletion of historical data [website]

What it doesn’t do (yet):

• Automated reports — the website explicitly says “Prisme doesn’t support automated report (yet) but you can easily build it yourself” [website]. That parenthetical “yet” is honest, but it’s a gap for teams that run weekly traffic reports.
• Mobile apps — web dashboard only.
• The AlternativeTo listing flags no mobile platform support [1].

---

Pricing: SaaS vs self-hosted math

Prisme Cloud:

• Free trial: 30 days, no credit card required [website]
• Paid plan: €8.99/month (listed as “limited time only”) — unlimited websites, unlimited pageviews/events, unlimited dashboards, GDPR compliance, built-in analytics dashboard [website][1]
• One plan. No tiers. No per-site fees. No pageview caps.

Self-hosted (Open Source):

• License: €0 (AGPL-3.0) [README][3]
• VPS to run it: ~$5–10/month (Hetzner, Contabo, DigitalOcean)
• Note: you’re running both Prisme and a Grafana instance, which increases RAM requirements

Competitor comparison at 100K pageviews/month:

• Plausible Cloud: ~$19/mo (100K pageview tier, scales to $69+ for 1M)
• Fathom: ~$15/mo (100K pageview tier)
• Matomo Cloud: ~$23/mo
• Google Analytics: free (you pay with data sovereignty)
• Prisme Cloud: €8.99/mo flat, regardless of pageview count

Self-hosted savings math:

For a site doing 500K pageviews/month, Plausible Cloud is $39/mo ($468/year). Fathom is $44/mo ($528/year). Prisme Cloud is €8.99/mo (~$118/year). Self-hosted Prisme on a $7 Hetzner VPS is $84/year — plus your time to set it up and maintain it.

The flat-price model is Prisme’s most concrete commercial differentiator. The catch: the “limited time only” label on the €8.99 price suggests this could change [website]. No pricing history is available to judge how stable it’s been.

---

Deployment reality check

The README’s installation path is Docker-based, with a guide for self-hosting via Docker Compose [README]. The actual deployment is more involved than the “3-minute setup” marketing suggests, because you’re deploying two services: the Prisme analytics server and Grafana.

What you actually need:

• A Linux VPS (minimum 2GB RAM; 4GB recommended if running Grafana and Prisme alongside other services)
• Docker and docker-compose
• A domain name with HTTPS (Caddy or nginx as reverse proxy)
• Grafana instance (bundled or separate)
• PostgreSQL or ClickHouse as the event store (the README and Docker image suggest it bundles its storage)

What can go sideways:

• The Grafana dependency means two systems to update, monitor, and secure — not just one. If Grafana has a CVE, you need to patch it. If Grafana’s API changes, Prisme’s dashboard integration may break.
• The project is maintained by a solo developer with 123 GitHub stars. This is a concentration risk. If Alexandre stops maintaining it, you own the Go codebase.
• No documented upgrade path or migration guides are visible in the available materials. For a production analytics stack, this matters.
• The AGPL-3.0 license is a hard stop if you’re planning to embed Prisme in a commercial product or white-label it for clients [3].

Realistic time estimate for someone comfortable with Docker: 1–2 hours for a working instance including Grafana setup and HTTPS. For someone new to self-hosting: 4–6 hours including debugging and domain configuration.

---

Pros and Cons

Pros

• Flat unlimited pricing. €8.99/mo for unlimited everything is a strong value proposition versus per-pageview competitors at scale [website][pricing page].
• Genuinely cookie-less. GDPR, PECR, PPCA, and Schrems II compliant by architecture — not by configuration [README][3]. No consent banner required. Your legal team will sign off quickly.
• Lightweight tracking script. 2kB versus Google Analytics’ ~45kB bloat. The noscript pixel path at 35 bytes is a nice touch for accessibility and tracking completeness [README].
• Grafana dashboards. Custom panels, filters, user and team management, multi-org support — you get a mature BI layer without the Prisme team having to build all of it [README].
• Real-time data. No 24–48 hour processing lag like Google Analytics [3][website].
• SQL API access. Direct query access to your events data without going through aggregated views [website].
• High throughput. 50,000+ req/s on consumer hardware is a credible performance claim backed by published k6 test output in the README [README].
• Forever data retention. No automatic deletion of historical pageview data [website].
• EU hosted. For European businesses, cloud data stays in the EU [website].

Cons

• Early-stage with thin community. 123 GitHub stars, 2 forks, no independent blog reviews found — this is genuinely early software [1][README]. The risk is maintenance continuity.
• AGPL-3.0 blocks commercial embedding. Unlike Plausible (also AGPL) or MIT-licensed alternatives like Vince, AGPL means any commercial product incorporating Prisme must open-source itself [3]. This rules it out for building analytics into your own SaaS.
• Grafana dependency adds operational weight. You’re not deploying one service — you’re deploying two. Grafana has its own updates, security patches, and configuration surface [README].
• No automated reports. The website explicitly acknowledges this gap [website]. If you need scheduled weekly email digests of traffic data, you build that yourself.
• “Limited time only” pricing is vague. The €8.99/mo price carries that label with no stated expiration or price increase schedule [website]. Unclear what happens to that price as the product matures.
• Solo maintainer. One developer, no visible team page, no company backing mentioned. This is a risk for any tool you’re betting your analytics stack on long-term.
• No Trustpilot, G2, or Capterra reviews. There’s no independent user review corpus to draw on. The AlternativeTo page has zero comments or user reviews [1].

---

Who should use this / who shouldn’t

Use Prisme if:

• You’re paying $15–$40/mo for Plausible or Fathom and your traffic is growing — the flat pricing makes Prisme cheaper past certain volume thresholds.
• You need GDPR/Schrems II compliance with zero configuration and no cookie banner, and you want to audit the code yourself.
• You already run Grafana internally and want to add analytics as a dashboard alongside your other metrics.
• You’re a technical founder comfortable with Docker who wants to own your data completely for $5–10/mo in VPS costs.
• You’re building on a small budget and the €8.99/mo flat plan is attractive compared to per-tier competitors.

Skip it if:

• You’re not technical and can’t debug a Docker Compose setup. Use Plausible Cloud or Fathom — they’re managed, polished, and their support teams exist.
• You want to embed analytics into your own commercial product. AGPL-3.0 means you’d have to open-source your product [3].
• You need automated scheduled reports today. That feature doesn’t exist yet [website].
• You’re evaluating tools for a 50+ person organization that needs audit logs, SSO, and enterprise support. The solo-maintainer model is the wrong bet at that scale.
• You want a large community, plugin ecosystem, and years of production war stories. Use Matomo.

Skip it (stay on Google Analytics) if:

• You genuinely don’t care about visitor data sovereignty, your site traffic is modest, and free is your hard constraint. GA4 is free and Google isn’t going anywhere.

---

Alternatives worth considering

From the AlternativeTo listing and the web analytics category:

• Plausible Analytics — the most direct comparison. Also AGPL-3.0, also cookie-less, also privacy-first, but with 20K+ GitHub stars, a larger community, a polished managed cloud, and per-pageview pricing that climbs as you grow [2]. More mature. More expensive at scale.
• Matomo — the most feature-complete open-source analytics platform. GPL-3.0. Much larger ecosystem, plugin store, extensive integrations. Self-hosted is free. More complex to run than Prisme [2].
• GoatCounter — minimal, privacy-friendly, open source. Free for non-commercial use. Simpler than Prisme, no custom events, very light [2]. Good for personal projects.
• Vince — AGPL-3.0, self-hosted, zero dependencies, automatic TLS. Newer project but similar positioning [2]. Alternative to consider if you want to avoid the Grafana dependency.
• Rybbit — also AGPL-3.0 open source, newer entry with 26 likes on AlternativeTo, real-time dashboards, multi-site support [2].
• Fathom Analytics — proprietary SaaS, no self-hosted option, privacy-first, polished product. Starts at $15/mo. Good if you want zero operational overhead and don’t mind closed source.
• Google Analytics — free, ubiquitous, powerful, privacy nightmare. Your data is Google’s product.

For a non-technical founder escaping Google Analytics, the realistic shortlist is Plausible vs Prisme. Pick Plausible if you want a mature product with a real support team and don’t mind per-pageview pricing. Pick Prisme if you want flat pricing, Grafana-based custom dashboards, and you’re comfortable owning the infrastructure.

---

Bottom line

Prisme Analytics makes a coherent pitch: flat-price, cookie-less, GDPR-compliant web analytics with Grafana-backed custom dashboards and a performance profile that won’t crack under traffic spikes. The pricing math is genuinely favorable against Plausible and Fathom at high pageview volumes, and the technical architecture is sound — 50K req/s on consumer hardware, a 2kB script, direct SQL access to your events. These are real advantages.

The honest counterweight is that this is a 123-star project maintained by one developer, with no independent user reviews, a Grafana dependency that adds deployment weight, and no automated reporting. For a founder building on a tight budget who’s comfortable with Docker and wants to stop sending traffic data to Google, it’s worth a trial — the 30-day free tier with no credit card is a low-commitment entry point. For anyone who needs a production-grade analytics stack with long-term maintenance guarantees, Matomo or Plausible are the safer bets until Prisme’s community grows.

If the deployment is the blocker, that’s exactly what upready.dev handles for clients. One-time setup, you own the infrastructure, no recurring analytics bill beyond the cost of your VPS.

---

Sources

1. AlternativeTo — Prisme Analytics: About (11 likes, 125 stars, AGPL-3.0). https://alternativeto.net/software/prisme-analytics/about/
2. AlternativeTo — Prisme Analytics: Alternatives (top alternatives: Rybbit, Matomo, Google Analytics, Plausible). https://alternativeto.net/software/prisme-analytics/
3. Prisme Analytics official — “Prisme: Open Source Google Analytics alternative”. https://www.prismeanalytics.com/is/open-source-website-analytics/

Primary sources:

• GitHub repository and README: https://github.com/prismelabs/analytics (~123 stars, AGPL-3.0, written in Go)
• Official website: https://www.prismeanalytics.com
• Pricing page: https://www.prismeanalytics.com/pricing