Product analytics stack
For PMs, founders, or growth folks stacking 3 analytics tools when one platform covers events, sessions, and funnels.
The stack map.
A stack is not read tool by tool. It is read as work blocks: produce, validate, deliver, get paid.
Measurement & support
Measure, track, and understand what actually works.
Core: Plausible Analytics. Add-ons based on your actual use.
Other steps
Useful tools that complete the workflow.
Core: PostHog + Metabase. Add-ons based on your actual use.
A budget that stays healthy.
Frequently asked questions.
Did you define your 3 to 5 critical events before instrumenting?
No → you will end up with 50 custom events of which 45 you never look at. The only events that matter at early-stage: activation, D7 retention, the event before churn. Start only with those.
Do you look at your analytics more than once a week?
No → you are paying for reassurance, not for action. Set up 3 metrics max on a dashboard + one alert on anomalies (drop >20%). More frequent = noise, not signal.
Do you have a SQL database and want custom analyses?
Yes → open-source Metabase deployed on a VPS (€10-15/month) is structurally superior to any analytics SaaS for custom SQL queries. PostHog SQL Insights covers intermediate cases.
If this stack does not quite match your use case.