PostHog is an open-source product analytics platform founded in 2020 in San Francisco, USA, combining product analytics, session recording, feature flags, and A/B testing in one platform, with free self-hosting and cloud hosting, and 30,000+ GitHub stars.

Strengths and Features

✅ Founded in 2020 in San Francisco, USA, by James Hawkins and Tim Glaser (Y Combinator W20), fully open source under the MIT license
✅ Combines 6 capabilities in one platform: product analytics, session recording, feature flags, A/B testing, surveys, and data pipelines
✅ Cloud Free plan includes 10,000 events per month; Growth is usage-based at roughly $0.00025 per event (reference)
✅ Self-hosting is completely free with no event limits, 100% data ownership, and 30,000+ GitHub stars
✅ Offers 10+ mainstream SDKs (JavaScript, React, iOS, Android, Python, Node.js, etc.) plus a full REST API
⚠️ Cloud is billed per event, so costs can rise quickly for high-traffic products
⚠️ Self-hosting requires Docker/ClickHouse administration skills and frequent upgrades
⚠️ High feature density with a steeper learning curve for teams that only need basic web stats
⚠️ AI-driven insights are relatively limited, so deep smart analysis needs additional tools

Product Information

Category Analytics

Tags self-hosted A/B Testing Event Tracking Open Source Analytics Product Analytics Data Pipeline Session Recording Feature Flags PostHog

Company posthog

Service Region United States

Official Website https://posthog.com/

Recommendation Score 5 pts

Published 2026-08-09

Detailed Notes

Overview

PostHog was founded in 2020 in San Francisco, USA, by James Hawkins, Tim Glaser, and co-founders (Y Combinator W20), as a fully open-source product analytics platform under the MIT license. Unlike tools focused on website traffic, PostHog targets SaaS product teams, combining event tracking, funnel/retention/path analysis, session recording, feature flags, A/B testing, and data pipelines in a single platform. Its core promise is modern product analytics infrastructure that is open source, developer-friendly, and data-sovereign.

PostHog offers two modes: free open-source self-hosting (no event limits, data stays 100% on your own infrastructure) and PostHog Cloud (usage-based pricing, with 10,000 events per month on the Free plan). The project has 30,000+ GitHub stars, raised a $15M Series B in 2022 at a valuation above $1B, and is used by companies such as Airbus, Hasura, and Vendr, making it one of the most active open-source projects in product analytics.

Key Strengths

Fully open source and self-hostable: MIT-licensed and publicly auditable, with free self-hosting, no feature gating, and no event limits. Data stays 100% under your control, which is almost a must-have for finance, healthcare, and other compliance-sensitive industries.
Six capabilities in one platform: Product analytics, session recording, feature flags, A/B testing, surveys, and data pipelines combined, reducing multi-tool switching and duplicate instrumentation, with recordings, events, and flags naturally linked.
Excellent developer experience: 10+ mainstream SDKs (JavaScript, React, iOS, Android, Python, Node.js, PHP, Go) and a full REST API, autocapture to lower instrumentation effort, and Data Pipelines that export events to Snowflake, BigQuery, and Amazon S3.
Long-term cost control: Self-hosting has no event limits and a 2 vCPU / 4GB VPS can handle hundreds of thousands of daily events; Cloud Growth is billed at roughly $0.00025 per event (reference), staying competitive for small and mid-size volumes.
Privacy and compliance friendly: IP anonymization, data retention policies, user data export/delete, and privacy masking, with no data leaving your infrastructure when self-hosted; the cloud offers a DPA and GDPR/CCPA alignment - see the GDPR compliance checklist.

Product Ecosystem

Product Analytics

Event tracking (autocapture plus custom events), funnel analysis, retention analysis, trend analysis, and path analysis, with multi-dimensional breakdowns by source, device, and user attributes to locate conversion bottlenecks and stickiness.

Session Recording

Records real user sessions with an event timeline, browser console logs, network requests, and Core Web Vitals metrics, plus CSS-selector privacy masking and sampling rate configuration, so experience issues are found without switching tools.

Feature Flags and Experiments

Feature flags support progressive rollouts by percentage or attribute and remote configuration; the built-in A/B experiment engine uses Bayesian statistics for significance and confidence intervals, closing the ship-observe-decide loop.

Surveys and Feedback

In-product surveys target users by attributes and behavior, linking qualitative feedback with behavioral data to support product validation and user testing.

Data Pipelines and Integrations

Data Pipelines export events in real time to Snowflake, BigQuery, and Amazon S3, with webhooks to Slack and Discord; integrates with customer data platforms such as Segment and RudderStack - see the website API integration basics.

Deployment Modes

Self-hosting deploys via Docker Compose (PostHog + ClickHouse + PostgreSQL + Redis) in about 20-30 minutes; cloud hosting is ready to use with vendor-operated upgrades - see the Docker production deployment guide.

Limitations

Usage-based cloud billing: Billed per event, so costs fluctuate noticeably during traffic spikes; high-traffic products should evaluate self-hosting to control total cost of ownership.
Self-hosting needs ops skills: You maintain ClickHouse, PostgreSQL, and frequent version upgrades; teams without dedicated ops face a real learning curve.
High feature density: The broad feature set and many configuration options can be overwhelming for teams that only need basic web stats - consider a lighter tool such as Plausible.
Limited AI capabilities: AI insights focus on basic anomaly flags and trend forecasting; deep smart analysis requires pairing with other platforms.
Cloud is hosted in the US/EU: Users in mainland China may see higher reporting latency with the cloud; self-hosting on a local server improves collection stability.

Use Cases

SaaS product teams (Recommended ★★★★★): Event analytics, session recording, feature flags, and A/B testing in one place accelerate iteration; see OKRs and product metrics.
Data-sovereignty-sensitive industries (Recommended ★★★★★): Self-hosting keeps data on your infrastructure, satisfying finance, healthcare, and government compliance.
Teams that need self-hosting (Recommended ★★★★★): Open source, free, no event limits, with full data ownership.
Technical, developer-driven teams (Recommended ★★★★★): Rich SDKs, a complete API, and CI/CD-friendly workflows with low integration cost.
Mobile app teams (Recommended ★★★★): iOS, Android, Flutter, and React Native SDK coverage.
Site owners needing basic traffic stats (Recommended ★★): More about product behavior than web traffic - consider Plausible or Matomo.

Pricing

Plan	Monthly Events	Reference Price	Data Retention	Notes
Self-hosted (open source)	Unlimited	Free (server cost only)	Your choice	All features, data owned
Cloud Free	10,000 events	Free	7 days	Core features, community support
Cloud Growth	Usage-based	Roughly $0.00025/event (reference)	30+ days	All features, Slack support
Cloud Enterprise	Custom	Custom quote	Custom	SSO, audit logs, SLA, dedicated support

Cloud prices are public reference figures; the Growth rate is on the order of $0.00025 per event and Free includes 10,000 events per month - confirm exact values on the official website. Self-hosting is free and requires only server cost (a 2 vCPU / 4GB VPS is roughly $10-20/month).

FAQ

How do PostHog and Amplitude/Mixpanel compare? PostHog is open source, self-hostable, and all-in-one (including session recording and feature flags), while Amplitude and Mixpanel are closed-source SaaS focused on event analytics; choose PostHog for data sovereignty and cost - see the website analytics setup guide.
Is PostHog self-hosting really free? Yes. The self-hosted edition is fully free under the MIT license with no event limits; you only pay for servers and ops, and deployment takes about 20-30 minutes - see the Docker production deployment guide.
Can PostHog replace Google Analytics? They serve different purposes: PostHog focuses on product behavior analytics (events, funnels, retention, recording, experiments) while GA focuses on web traffic and ad attribution; for basic site traffic stats use other tools - see web analytics tools; see website analytics setup guide.
Is PostHog data secure? When self-hosted, data lives on your own servers with IP anonymization and privacy masking; for European users, GDPR requirements apply - see the GDPR compliance checklist.
How do I export data from PostHog? Data Pipelines stream raw events to Snowflake, BigQuery, or Amazon S3 in real time, and the REST API supports query and bulk import - see the website API integration basics.

Visit PostHog Website →