Reviewed by Harsh Desai · Last reviewed:
PostHog
Open-source product analytics with session replay, flags, and surveys
Best for
PostHog is an open-source, all-in-one product analytics platform from the PostHog team. It bundles event analytics, session replay, feature flags, A/B testing, surveys, heatmaps, error tracking, and a data warehouse layer into a single MIT-licensed toolkit. You can run PostHog as a managed cloud service or fully self-host it under MIT, and pricing is usage-based per product rather than per seat, so adding engineers, PMs, or analysts does not raise the bill.
What PostHog does:
- •Product analytics event tracking, funnels, retention, cohorts, and trends across web and mobile, with HogQL (their SQL flavour) for ad-hoc analysis when the visual builder runs out of road.
- •Session replay watch real user sessions to spot rage-clicks, dead-ends, and broken flows; integrates inline with funnel drops to jump straight to the failing replays.
- •Feature flags and remote config ship code behind flags, target by cohort or percentage rollout, and flip features without redeploying.
- •A/B testing and experiments statistical experiment runner that ties experiment exposure to product analytics events, with confidence intervals and recommended sample sizes built in.
- •Surveys and feedback in-product NPS, CSAT, and free-text surveys triggered on user attributes, with results piped into the same warehouse as the rest of your data.
- •Heatmaps and dashboards click and scroll heatmaps for any URL plus a flexible dashboarding layer that supports shared team boards.
- •Error tracking frontend and backend exception capture with stack traces and user context, replacing a separate Sentry-style tool for many small teams.
- •Data warehouse layer query data alongside your Stripe, Hubspot, or Postgres rows so revenue lives in the same place as product events for unified analysis.
- •Open-source under MIT the entire repo is MIT-licensed on GitHub, so you can self-host on your own Kubernetes, audit the code, or fork it without licence anxiety.
- •API, CLI, and SDKs first-party SDKs for JavaScript, Python, Ruby, Node, Go, Android, iOS, Flutter, and React Native, a
posthogCLI for project setup and CI flag toggles, plus a clean REST API for any language not yet covered.
Pricing:
- •Free $0/mo: 1M product analytics events, 5K session replays, 1M feature flag requests, 100K error logs, 1.5K survey responses; unlimited team members and a credit-card-free signup.
- •Pay-as-you-go from $0.00005/event/mo: usage-based pricing per product with tiered step-down rates as volume grows; billing caps available to prevent surprise spend.
- •Self-hosted $0/mo: free forever under MIT, you only pay for the cloud or hardware that runs PostHog. Suits regulated industries that need full data residency.
Limitations:
- •Self-hosting is a real ops job ClickHouse, Kafka, Postgres, and Redis make for a non-trivial deployment, so most teams under 20 engineers stick to PostHog Cloud rather than running their own cluster.
- •The UI can feel busy with nine products in one app, navigation has a learning curve, and new users sometimes need a day to find the right tool for a given question.
- •High-cardinality event volumes can spike costs usage-based pricing rewards careful event design; teams that send every user interaction without filtering can outgrow the cheap tiers quickly.
Our Verdict
PostHog scores 8.3/10 because it consolidates nine separate analytics tools into one open-source platform with a genuinely generous free tier (1 million events per month) and pricing that does not tax you per seat. The all-in-one architecture replaces Mixpanel plus Hotjar plus LaunchDarkly plus Sentry-lite plus Typeform-for-surveys with a single contract, which is a meaningful cost and integration win for most product teams in 2026.
For the Vibe Builder, PostHog is the cleanest path to grown-up product analytics without engineering a stack from scratch. Drop a snippet on your Next.js site, ship a feature flag, run a survey, and review session replays of real users; that is a one-afternoon setup that would otherwise need three or four tools wired together. The free tier covers most early-stage products end-to-end, and HogQL gives you SQL when the visual UI runs out.
For the Developer, PostHog is the rare analytics platform you can actually self-host under MIT and treat as part of your infrastructure. The SDKs cover every common language and runtime, the REST API is well-documented, and the data warehouse layer means you can join product events with Stripe revenue or Postgres rows without exporting to a separate destination. Self-hosting is a real ops job (ClickHouse, Kafka, Postgres, Redis) but the option is genuinely there.
Skip it if you only need lightweight web analytics; in that case, consider Plausible (lighter, EU-friendly, privacy-first) or try Amplitude when you want a polished product-analytics-only UI and do not want to manage events at the engineer level. Choose PostHog when product analytics, replay, flags, and experimentation should live in one self-hostable, MIT-licensed stack.
Related Tools
View allCompare PostHog With
Also Useful For
Frequently Asked Questions
How much does PostHog cost in 2026?
PostHog is free for the first 1 million product analytics events per month, plus 5,000 session replays, 1 million feature flag requests, 100,000 error logs, and 1,500 survey responses. Beyond that, you pay usage-based rates starting at $0.00005 per event with tiered step-down pricing, and self-hosting under MIT is free forever.
Is PostHog actually open source and self-hostable?
Yes. The entire PostHog repo is MIT-licensed on GitHub and you can self-host the full stack including ClickHouse, Kafka, Postgres, and Redis on your own Kubernetes cluster. PostHog Cloud is the managed alternative for teams that do not want to operate the infrastructure themselves; both offer the same product features.
PostHog vs Mixpanel: which product analytics tool should I pick?
Choose PostHog when you want product analytics plus session replay, feature flags, A/B testing, and surveys in one platform with optional self-hosting. Choose Mixpanel when you want a pure product-analytics tool with a polished, focused UI and do not need replay or feature flags. PostHog wins on breadth and price; Mixpanel wins on focus.
What languages and frameworks does PostHog support?
PostHog ships first-party SDKs for JavaScript, Python, Ruby, Node, Go, Android, iOS, Flutter, and React Native. It also exposes a REST API and HogQL (a SQL flavour) so any language not on that list can still send events. Server-side, edge, and cron-driven event sources are all common patterns.
Can PostHog replace Hotjar, LaunchDarkly, and Mixpanel together?
For most small teams, yes. PostHog session replay covers the Hotjar use case, feature flags and experiments cover LaunchDarkly, and product analytics covers Mixpanel. Heavy enterprise users (high-cardinality cohorts, complex experiment governance) sometimes still pair PostHog with a specialist tool, but the consolidation saves cost and integration overhead in 2026.
What is PostHog?
PostHog is Open-source product analytics with session replay, flags, and surveys.
Is PostHog free?
Yes, PostHog offers a free version.
Who should use PostHog?
PostHog is built for vibe builders who want AI to handle the technical work and developers looking to accelerate their workflow. Common use cases include product-analytics, session-replay, feature-flags, ab-testing, user-surveys, self-hosted-analytics.
What are the best alternatives to PostHog?
Popular alternatives to PostHog include Hubspot, Attio. Compare features and pricing in our Growth directory to compare options.
Affiliate link: we may earn a commission. How this works.
PostHog
Free tier available