Countly
Countly is an open-source product-analytics platform covering web and mobile, available as a self-hosted deployment as well as managed hosting. It collects events, sessions, and user properties through SDKs, with plugin-based features. Self-hosting means the event data can stay in infrastructure you control, which is its main posture distinction.
What this means
Countly is built around product-analytics primitives: events (actions users take), sessions, and user properties, captured via web and mobile SDKs. A plugin architecture extends it with additional features, and it spans both web and app measurement under one model.
It is positioned as product analytics rather than purely page-counting web analytics, focusing on user behavior and flows.
Self-hosting and data ownership
A defining option is self-hosting: you can run Countly on your own infrastructure (an open-source/community edition exists alongside managed offerings), which keeps event data within systems you control. That appeals to teams with data-residency or ownership requirements.
Self-hosting shifts operational responsibility to you and does not remove consent obligations. Confirm current editions, plugins, and feature specifics against the project documentation.
- Events, sessions, user properties via SDKs
- Web and mobile under one product-analytics model
- Plugin-based feature extension
- Self-hosting option keeps data in your infrastructure
How it appears in analytics and logs
Countly data is event- and session-based product analytics. When self-hosted, the data lives in your infrastructure, so collection completeness and bot handling depend on your deployment and SDK setup.
Diagnostic use case
Use Countly when you want product analytics across web and mobile with the option to self-host, keeping event data in infrastructure you operate rather than a vendor cloud.
What WebmasterID can help detect
Countly's product-analytics events sit alongside WebmasterID's first-party traffic intelligence; bot separation remains a distinct concern that product SDKs do not inherently solve.
Common mistakes
- Assuming self-hosting alone satisfies privacy obligations.
- Underestimating operating a self-hosted deployment.
- Expecting bot filtering that product SDKs do not provide.
Privacy and accuracy notes
Self-hosting can keep event data within your own infrastructure, but it does not by itself satisfy consent or minimization obligations — configuration and SDK setup decide what is collected. This is educational, not legal advice.
Related pages
- PostHog: product analytics plus more
PostHog is an open-source platform that bundles product analytics (events, funnels, retention) with adjacent tools such as session replay, feature flags, and experiments. It can be self-hosted or used as a hosted cloud service. Like other event-centric tools, its analytics depend on the events you instrument, while the broader suite aims to keep several product-engineering tools in one place.
- Self-hosted vs cloud analytics
Choosing between self-hosted and cloud (vendor-hosted) analytics is mainly a trade-off between data ownership and operational effort. Self-hosting keeps raw data in your own database and gives you control over retention, but you run, secure, and update the software. Cloud is operated for you but the data lives with the vendor. Neither is universally better.
- Product analytics vs web analytics
Product analytics and web analytics are different categories that are easy to conflate. Web analytics centers on pages, sessions, and acquisition sources; product analytics centers on events, users, and in-product behavior such as funnels and retention. Neither replaces the other — they answer different questions, and many teams use both.
- Events docs
Designing events for product analytics.
Sources and verification notes
- Countly — Documentation / developer resourcesPublic docs describe events/sessions and self-hosting; confirm current editions and features against the docs.
Last reviewed 2026-06-24. Facts are checked against primary/official sources where available; uncertain specifics are marked “Data not yet verified” rather than guessed.