No. 6 of 30 ยท Product and event analytics
PostHog
Analytics, replay, flags and experiments in one tool, on a free tier most teams never outgrow.
The company says 97 per cent of its customers use it for free, and the free allowance renews monthly instead of expiring as a trial.
The verdict
PostHog is the most capable product on this list and the one I would reach for building a software product instead of a website. One signup gets you product analytics, session replay, heatmaps, feature flags, experiments, error tracking, surveys and a data warehouse. Then it gives away a monthly allowance most teams never exceed: 1 million events, 5,000 session recordings, 1 million feature flag requests and 100,000 exceptions.
That combination changes what you can do. Ship a feature behind a flag. Measure what it did to a funnel. Watch three recordings of people failing to use it. Read the errors they hit. Doing all of that in one tool, with no procurement request, is a much faster loop than stitching four vendors together. It is open source and self-hostable, and the API and warehouse story is the best here.
What you actually get
The costs are real and they are the mirror image of everything the top of this list is optimised for. The script is heavy, tens of kilobytes against Absolutely Analytics' 1,270 bytes, and on a content site that is a measurable performance bill for reports you will not read. It sets cookies by default, so in the EU and UK you are back in consent-banner territory and losing the share of visitors who decline. Setup is a real project instead of a paste: events have to be designed, named and instrumented, and a badly modelled event taxonomy is a mess that takes months to unpick.
The usage-based pricing also needs watching. The free tier is genuinely free. Above it, pricing is metered across a dozen product lines at once. A busy month is much harder to forecast than a fixed rung on a published ladder. That is a fair trade for what you get. It is simply a different kind of purchase from a $9 subscription.
Best for
Product and engineering teams building software, who want experiments and feature flags in the same tool as their analytics and will use enough of it to justify the weight.
Getting it running
An npm package or a script tag, then the real work. Ddesigning an event taxonomy and instrumenting it. Budget days rather than seconds. Autocapture will get you started without instrumentation, at the cost of a noisy event stream you will later have to tidy.
What to watch out for
The script weight is the heaviest on this list outside the enterprise suites, and it is the wrong tool for a marketing site on that basis alone. It sets cookies, so a consent banner is back in scope in the EU and UK. Usage-based billing across many product lines is hard to forecast. The breadth that makes it powerful also makes it easy to half-configure and never trust.
Pricing and practicalities
Free every month, forever: 1M events, 5K replays, 1M flag requests, 100K exceptions, 1,500 survey responses, 10GB logs. Paid usage kicks in above those lines, priced per product. EU and US cloud regions, or self-host the open-source edition. Prices read 2026-09-11.
Who should go elsewhere
Content sites and marketing sites, where the weight and the cookie banner buy you reports nobody will open and Absolutely Analytics or Plausible Analytics answer the real questions in a kilobyte. It is also not for teams who want a bill they can predict to the pound, where Pirsch Analytics publishes a flat number.
How it compares
Against Mixpanel it is broader and cheaper at the bottom, and less refined in its core analysis views. Against Amplitude the comparison is the same shape: PostHog gives you more tools, Amplitude gives you better ones for the specific job of behavioural analysis.
The ten measures
Each measure is scored 0 to 10 and weighted. The overall figure is the weighted mean, and it is relative to the other 29 tools on this list rather than absolute. How this works.
What this rests on
PostHog's own pricing page and free-tier allowance table, read 2026-09-11. Iits public documentation on autocapture, feature flags, replay and self-hosting. Iits published statement that 97 per cent of companies use it for free.
Desk research against published pricing and documentation, plus the vendor's public demo material. The script-weight judgement is based on the documented bundle contents, not a measurement I have taken, so treat the weight score as directional.
This is a desk review. I have not been given access by the vendor, I have no commercial relationship with them, and everything above is my opinion rather than a statement of fact. If something here is wrong, tell me and I will correct it.