PostHog
Why PostHog
PostHog is an open-source product analytics platform that bundles funnels, retention and trends with session replay, feature flags, A/B experiments, surveys and a data warehouse. It suits B2B product and engineering teams who want to understand in-product behaviour, such as where new accounts drop off in onboarding. It is developer-focused and can run on PostHog Cloud or on your own infrastructure.
What it does
PostHog is a product analytics platform that bundles many jobs. The core is event-based analytics with funnels, retention, paths and trends. On top of that it offers session replay, error tracking, web analytics, feature flags, experiments, surveys, a managed data warehouse, a customer data platform and workflows. It can pull in data from more than 800 other sources, and it has an AI assistant called PostHog AI.
It is open source. You can use PostHog Cloud or run it on your own infrastructure.
Why you would need it
You need it when you have questions about what people do inside your product. Where do new accounts drop off in onboarding? Which feature keeps customers around? Did the change you shipped last week move activation? Web traffic tools do not answer those.
Instead of five tools, you watch the funnel, replay the session that explains the drop, and roll out a fix behind a flag from the same place.
Where it fits
PostHog sits next to your app. You install an SDK, send events, and the rest builds on that data. It feeds product decisions, experiments and engineering fixes. It replaces separate analytics, replay, flag and survey tools.
It is not a web traffic tool for marketing. For that job, compare Google Analytics. Teams comparing product analytics alone can also look at Mixpanel, Amplitude and Heap.
What stands out
- Many products, one platform. Analytics, replay, error tracking, flags, experiments and surveys share the same data.
- A free tier on every product. The pricing is usage-based, with a free allowance for each product and no card needed.
- Open source with self-hosting. That helps when you have data residency or privacy constraints.
- A free hosted MCP server. Your coding agent can query PostHog, ship a flag or read a stack trace from your editor.
My take
The strength is consolidation. One source of truth with less glue between vendors is worth a lot, and being able to drop into SQL when the built-in reports run out is a real advantage.
The cost is breadth. There is a learning curve, and the surface area can feel like a lot if you only wanted one module. Good answers depend on good event design, so messy tracking gives messy analytics. Most teams should run the cloud version, since self-hosting is real work.
Verdict
Pick PostHog when you build a software product, have at least some engineering capacity, and want analytics, replay and flags together. Skip it if you mainly want marketing traffic and channel reports, or if nobody can own the SDK install and the event plan.
Books behind it
Notes
Your note
Before you choose
It needs engineering time
PostHog rewards teams that can install the SDK and define events and properties. Non-technical marketing teams who want sessions by source will find it heavier than they need.
Read moreShow less
Breadth means a learning curve
There are many products in one place. Start with one or two, such as analytics and replay, and add others when you need them.
Usage-based costs need a limit
Pricing is based on usage per product, with a free allowance on each. Set billing limits and watch event volume, especially with replay and many users. Some AI features in the MCP server may be billed as AI spend.
Self-hosting is work
Running PostHog on your own servers gives control over your data but also gives you the maintenance. Most teams sensibly start on the cloud.
Alternatives to compare
For web behaviour, compare Hotjar and Microsoft Clarity. For errors, compare Sentry. For pure product analytics, compare Mixpanel and Amplitude.