The whole PostHog suite

You don't need all of it. For this course you'll mainly use Product Analytics, Web Analytics, and Session Replay, plus Data Pipelines later for sending conversions to ad platforms. Here's the whole suite so you know what's there, and what each piece does for a growth team specifically:

• Product Analytics — event-based analysis: trends, funnels, retention, paths, stickiness, lifecycle, and raw SQL (docs). Includes autocapture (it records clicks and pageviews with no code). This is the core of the course — it's where you find the leaks.
• Web Analytics — a Google-Analytics-style dashboard (visitors, sessions, bounce rate, channels, geography) sitting on the same data, with far less setup. The marketing-site view, for the acquisition numbers.
• Session Replay — watch real user sessions play back, with console logs and network requests captured (docs). This is how you find the why behind a drop-off — the qualitative layer under the quantitative funnel. Five minutes of replay beats an hour of staring at a chart.
• Feature Flags — turn features on/off for a percentage of users or a targeted cohort (a saved group of users defined by shared traits or behaviour — everyone on the paid plan, say), without deploying (docs). The "ship behind a flag" half of the growth-as-engineering loop.
• Experiments (A/B testing) — built on feature flags; measures whether a change actually moved a metric, with statistical significance (docs). This is how a hypothesis becomes a measured result rather than a hunch — and the instrument behind the experimentation program you'll run in the experiments chapter.
• Surveys — in-app surveys — NPS (Net Promoter Score) and open feedback — targeted at specific users. The cheapest way to get the why straight from the user.
• Error Tracking — captures exceptions, groups them into issues, and links each one to the session where it happened. Production health that's joined to behaviour.
• LLM Analytics — PostHog's headline 2026 product: traces of individual LLM calls (tokens, latency, cost), per-user cost attribution, and evals. Relevant if you ship AI features and need to see what they cost per user.
• Data Warehouse — connect external sources (Stripe, HubSpot, your database) and join them to your event data with SQL. This is how revenue from Stripe meets behaviour from your app without a separate ETL (extract, transform, load) tool.
• Data Pipelines (CDP, a customer data platform) — Sources (managed ingestion), Destinations (send your events out to Snowflake, ad platforms, Slack, a CRM), and Transformations (clean or enrich events in flight). You'll use a Destination in the ads chapter to send conversions to Meta.
• PostHog AI ("Max") — an in-product assistant that answers questions in plain language, writes SQL for you, and builds insights. Useful when you don't yet know the query language, which is most people on day one.

⚠️ PostHog also ships newer products — Logs, Workflows (its own messaging layer), and Inbox — that are evolving quickly. Treat their exact status and limits as a moving target and check the docs.

Of that list, two products are where a growth team actually lives. The next piece is three minutes of each one working.