PostHog: What It Actually Does, What It Costs, and Who It Fits
An open-source, developer-led product OS that bundles product analytics, feature flags, session replay, and A/B testing on one shared event pipeline, priced per-product on usage rather than per seat.
PostHog is a product analytics platform, not a web or marketing analytics tool like GA4: it tracks in-product user behavior through events, and then bundles that data with feature flags, session replay, and A/B testing under one roof instead of treating analytics as a standalone product. It's the stronger pick for a developer-led product team that wants to self-host or stay close to the code, pay only for what it actually uses across each product, and avoid stitching together four separate vendors, and a less natural fit for a team that just wants a polished, PM-facing analytics interface without engineering involvement.
What PostHog Is For
PostHog is a product analytics platform used to understand how people actually behave inside a website or app, not how they arrive at it. It's event-based like GA4, but where GA4 is built for marketers tracking web and app traffic for advertising and acquisition, PostHog is built for product and engineering teams tracking in-product behavior: feature adoption, activation, retention, and the specific paths users take toward or away from converting. What sets it apart from Amplitude or Mixpanel is scope: PostHog isn't packaged as an analytics tool with a few bonus features bolted on, it's positioned as a full 'product OS', combining analytics, feature flags, session replay, A/B testing and experimentation, surveys, and error tracking on one shared event pipeline, all built open-source-first with a self-hostable option alongside its managed cloud.
It's built for developer-led product teams that want to ship a feature behind a flag, watch how it performs through analytics and session replay, and run an experiment on it, all inside one tool and one data model, rather than exporting events to a separate flagging vendor and a separate testing vendor. The open-source core and self-hosting option also make it a natural fit for teams with data residency requirements or an engineering culture that prefers to own its own analytics infrastructure.
Key Features
- Product analytics (event tracking, funnels, retention, cohorts, and SQL-backed custom queries) built on the same pipeline as every other product
- Feature flags and A/B testing/experimentation bundled natively, so a flag rollout and its performance data live in one place instead of two vendors
- Session replay to watch real user sessions, billed per recording with a free monthly allowance rather than gated behind a separate product
- Open-source core with a self-hosting option, alongside a fully managed cloud version, for teams that want to control where event data lives
- Error tracking and surveys included as additional products on the same platform and pricing model
- Per-product, usage-based pricing so a team pays only for the events, replays, or flag requests it actually consumes, not a flat platform fee for features it doesn't use
Pros & Cons
Pros
- ✔ A genuinely all-in-one platform: analytics, feature flags, session replay, experimentation, surveys, and error tracking share one event pipeline instead of requiring separate tools
- ✔ Open-source with a self-hosting option, appealing to engineering-led teams that want infrastructure control or have data residency requirements
- ✔ A usable free tier across every product (1 million events, 5,000 session replays, and 1 million feature flag requests per month) with no credit card required
- ✔ Per-product usage-based pricing means a team is not paying for a bundle of features it does not use, unlike a flat per-tier platform fee
Cons
- ✕ The developer-first interface and setup assume some engineering involvement, a less polished, self-serve experience for a purely PM-facing team than Mixpanel
- ✕ Usage-based, per-product pricing across events, replays, and flag requests can be harder to forecast at scale than a single flat plan price
- ✕ Behavioral cohorting and pathfinding analysis are not as deep or as mature as Amplitude at a comparable spend, since PostHog spreads its engineering effort across more products
Pricing
PostHog has a genuinely usable free tier.
| Tier | Price | What you get |
|---|---|---|
| Free | $0 | 1 million events, 5,000 session replays, and 1 million feature flag requests per month, 1 project, roughly 1 year of data retention, unlimited team members |
| Pay-as-you-go | Starts around $0.00005 per event, $0.005 per session replay, and $0.0001 per feature flag request beyond free limits, with volume discounts as usage grows | Full platform access with up to 6 projects and roughly 7 years of data retention, billed per product based on what is actually used |
| Boost / Scale add-ons | Boost from roughly $250/month, Scale from roughly $750/month | Higher-volume platform packages with added support, retention, and query performance for teams past the pay-as-you-go entry tier |
| Enterprise | Custom-quoted | SSO, advanced permissions, dedicated support, and compliance features for larger organizations |
Every product (analytics, replay, flags, experiments, surveys, error tracking) is metered and billed separately, so a team's actual bill depends heavily on which products it uses and at what volume.
Pricing checked as of late 2026 and hedged as approximate - PostHog prices each product on its own usage-based ladder rather than a single flat tier, and per-unit rates, free allowances, and add-on packages change fairly often, so confirm current numbers directly with PostHog before budgeting.
Common questions
No. GA4 is a marketing and web analytics platform built for tracking acquisition, traffic, and ad conversion. PostHog is a product analytics platform built for tracking in-product user behavior, feature adoption, and retention, and it bundles that with feature flags, session replay, and experimentation, typically used by product and engineering teams rather than marketing teams.
Yes. PostHog's core platform is open-source, and it offers a self-hosting option alongside its fully managed cloud version, which appeals to engineering teams that want to control where event data lives or have specific data residency requirements.
PostHog bundles more into one platform: analytics, feature flags, session replay, A/B testing and experimentation, surveys, and error tracking all share one event pipeline and one usage-based pricing model, rather than analytics being the core product with a narrower set of bundled extras.
Yes. The free tier covers 1 million events, 5,000 session replays, and 1 million feature flag requests per month with no credit card required, enough for a small team to run analytics, flags, and replay together without paying anything.
It takes more work than a flat per-tier plan. Each product, events, session replays, feature flag requests, and more, is metered and billed on its own usage ladder, so the total bill depends on which products a team actually uses and at what volume, rather than one flat monthly number.
Not sure PostHog is the right product analytics fit?
We look at your actual event volume, which products you'd actually use (analytics, flags, replay, experiments), and whether self-hosting matters to your team before recommending a product analytics setup.
Ready to get started?
We usually reply within 24 hours.