Why PostHog vs the alternatives
A growth team should be able to say why it chose its tools. PostHog is this course's default but not the only honest choice — the real landscape (Mixpanel, Amplitude, GA4, Segment, Heap, Statsig, Plausible, build-your-own) and where PostHog wins and loses.
In this chapter's workbook
Everything you write, answer, and tick below saves to your personal workbook. Sign in free to keep it across visits.
A growth team should be able to say why it chose its tools, not just that it did. PostHog is the course's default, but it is not the only honest choice — here's the real landscape and where PostHog wins and loses.
- Mixpanel / Amplitude — the incumbents of product analytics, and genuinely excellent at the analytics itself (Amplitude's behavioural cohorts and Mixpanel's reports are mature; on several individual analysis features they remain ahead). What they're not is all-in-one: no session replay, no feature flags, no CDP, no error tracking in the same bill — you'd stitch those from other tools, each with its own identity model to reconcile (PostHog's own comparison is more even-handed than you'd expect from a vendor, and worth reading for the feature-by-feature truth). Choose them over PostHog if deep, polished analytics is the only thing you need and you already have replay/flags elsewhere. Choose PostHog if you want one pipeline under everything, which is most small teams.
- Google Analytics 4 (GA4) — free and ubiquitous, but it's web analytics, not product analytics: aggregated, sampled at scale, session-centric, and not built for user-level funnels or retention cohorts (the events-and-persons piece covered why the data model itself is the difference). It will tell you traffic by channel; it will not honestly answer "do users who create a project in week one retain better." Keeping GA4 around for marketing-channel reporting is fine — asking it to run your activation funnel is the part that won't work.
- Segment / RudderStack — these are a pipe, not an analytics tool. They collect events once and fan them out to many destinations (including PostHog). Complementary, not competitive — and for a small team, often overkill: PostHog's own CDP and a direct server-side capture usually cover it. Reach for a dedicated CDP when you have many destinations and want one canonical event schema feeding all of them. RudderStack is the warehouse-first, open-source-leaning option here.
- Heap — autocapture-first, like PostHog's autocapture taken as the whole product. Good if you never want to define events; weaker on the breadth PostHog gives you.
- Statsig / LaunchDarkly — best-in-class experimentation and flags, but single- purpose. If A/B testing is your core discipline they're stronger than PostHog's experiments; you'd pair them with an analytics tool.
- Plausible / Fathom — privacy-first, cookieless web analytics. Lovely for a content site or a landing page where all you want is traffic and referrers without a consent banner. They are not product analytics at all — no persons, no funnels, no retention — so they answer a different question than this course asks. Fine to keep one running on the marketing site alongside PostHog.
- The build-your-own stack — Snowflake/BigQuery + dbt + a BI tool (Looker, Metabase) fed by Segment. This is where bigger data teams end up, and it's genuinely more powerful and flexible than any packaged tool. It's also a team and a budget you don't have yet. The honest rule: if you already live in a warehouse-native stack, PostHog is a duplicate and you should send events to your warehouse instead. If you don't, PostHog gives you 80% of that stack's value this afternoon, for free.
The one-line summary: PostHog wins on "one event pipeline under everything plus a generous free tier," which is exactly the shape a small growth team needs. It loses to single-purpose tools on depth-in-one-area, and to a warehouse stack on ultimate flexibility. For the team this course is written for, the trade is clearly worth it.
Check-in
What's actually running on your product today?
A note for your kind of startup
The choice barely changes by business type — PostHog fits a marketplace, a consumer app, a dev tool, or an e-commerce brand equally — but the emphasis does. Consumer apps lean hardest on retention and replay. B2B/marketplaces care about Group Analytics (roll events up to the account/company), which is the one place to check pricing first. Dev tools get unusual value from autocapture on docs + the warehouse join to track the docs-to-signup path. E-commerce often keeps GA4 for channel reporting alongside PostHog for on-site behaviour. Pick the tool once; tune the emphasis to your funnel.
Before you move on
This chapter's workbook items — sign in free and they save as you go:
What PostHog costs
Usage-based and free until you're at real scale: every product has a monthly free tier that resets, you're billed once at capture (storing history is free), and roughly 90% of companies use it for free. The model, the tiers, and the two gotchas.
Set up your account
Four small things worth doing while this chapter is fresh — about ten minutes — then review the chapter's key ideas and check yourself with the quiz.