Fathom Analytics and Umami both make a credible case for privacy-conscious website measurement. The decision for a small website team is whether to buy a managed dashboard with little maintenance or run a broader analytics platform that can also be self-hosted. A third option, Clics, matters when the team wants the complete traffic-to-conversion picture and a separate record of verified AI crawler requests.
The comparison below uses the same requirements for all three: setup, acquisition reporting, events and revenue, goals and funnels, session detail, privacy, AI traffic, access outside the dashboard, and the work of keeping the system running. The choice depends on which requirement is firm.
Setup and operating cost
Fathom is a hosted product. Add its script, then read traffic and events in one dashboard. It provides email reports, shared dashboards, GA4 import, and a documented API. Fathom wins when the team wants a narrow reporting surface and does not want to administer an analytics server. Its paid plans are based on traffic volume; check the current pricing page for your volume before choosing it.
Umami gives you a managed Cloud option and an open-source self-hosted option. Cloud removes infrastructure work; self-hosting gives the team control of the deployment and database but adds upgrades, backups, and monitoring. Umami wins when deployment control is a requirement. Price comparisons must include that operating work as well as subscription fees.
Clics is managed and starts with a lightweight browser tracker. It keeps visitor, campaign, conversion, and session reporting in one product. Its pageviews and custom events are billed by monthly volume. Clics wins the setup decision for a team that wants deeper website reporting without managing a server.
Website analytics and conversions
Fathom reports visitors, pages, sources, campaigns, and real-time activity. Events can represent signups or purchases and carry monetary values. Saved views and scheduled emails suit a marketing team that revisits the same questions. Its focus is an advantage if those reports cover the job; the current feature documentation does not describe a built-in multistep funnel builder or a session replay workflow.
Umami v3 goes further than a minimalist dashboard. Its current documentation covers custom events, goals, ordered funnels, journeys, revenue, segments, retention, heatmaps, and opt-in session replays. Those tools fit teams investigating behavior and conversion paths. They also bring more configuration and privacy decisions, particularly if replay is enabled. Umami wins when replay or self-hosted behavioral analysis is mandatory.
Clics covers pages, referrers, UTMs, events, page and event goals, revenue, multistep funnels, and individual session timelines. A team can inspect which campaign brought people to a pricing page, where a signup funnel loses them, and what happened in a visit. The custom event API handles actions that pageviews cannot represent. Clics wins for a complete website analytics workflow when replay and heatmaps are not requirements.
Privacy and data control
All three describe cookieless measurement. Fathom is hosted and documents privacy handling and EU data isolation. Umami can be self-hosted, which changes who operates and stores the data. Clics is managed and documents cookie-free measurement. A cookie-free analytics script does not settle every legal question for the rest of a site; review the full stack and the vendor terms.
AI referrals
AI referrals are human visits whose browser sends a recognizable AI-tool referrer. Fathom's current site says teams can explore AI referrals and filter the dashboard to see what those visitors do. Give it credit for that; this is no longer a Clics-only distinction.
Clics has a dedicated AI Analytics view with provider tabs, landing pages, referrers and UTMs, and custom events from AI-referred sessions. Some Overview filters, including Goal, are unavailable inside it. Umami's current feature documentation supports referrers and conversion analysis, but I could not verify a dedicated AI-assistant report in the pages reviewed. For any of these tools, a direct visit or stripped referrer can escape AI attribution. None of this identifies an exact model version.
AI crawlers
AI crawlers are automated requests, not the people described above. Fathom says it filters bots from visitor reports. Umami's reviewed docs do not establish a comparable verified AI crawler report. Those statements do not prove either product lacks another way to inspect server logs.
Clics documents a separate server-side package. It records a supported crawler only when both the user agent and source IP match the operator's published identity. The dashboard can show provider and requested pages; response status is available when the integration sees the response, while request-only tracking cannot provide it. You need a production project, a project-scoped crawler token, and server integration in addition to the browser tracker. A crawler request proves neither indexing nor later use of the page. AI crawler tracking is free for all Clics users; crawler requests do not count as consumed events.
Access beyond the dashboard
Fathom documents an API and CSV exports. Umami documents an API and recently added a read-only Cloud MCP server. Clics documents a REST API, TypeScript SDK, CLI, and MCP for website analytics and workspace tasks. Agent access is therefore not exclusive to Clics. The practical difference is whether the team's preferred interface and the needed tool are supported. Clics also documents a dedicated AI crawler analytics endpoint alongside its general stats API.
Final verdict
Choose Fathom when a hosted, compact dashboard, scheduled reports, and event value tracking cover the whole job. Choose Umami when self-hosting, replay, heatmaps, or deeper behavior reports are requirements. For a team seeking complete managed website analytics without those dependencies, choose Clics: its documented goals, revenue, funnels, and session timelines cover the core work, while dedicated AI referral analysis and verified crawler reporting extend the same product. Plan the separate server-side crawler setup before treating that last capability as available.
Top comments (0)