Plausible vs Mixpanel: could you build either?

Why Plausible gets Build it

• A pageview is one beacon request incrementing a counter — no cookies, no user identity, no consent banner needed
• Aggregating by page, referrer, and country from request headers is simple grouped counting
• A dashboard with a line chart and a few top-N tables is standard reporting UI

Why Mixpanel gets Borderline

• Sending an event (user did X, with properties) to an ingestion endpoint is a simple, well-documented pattern
• Basic counts and time-series charts over events are standard aggregation queries
• For a smaller product with modest event volume, an ordinary database handles this fine

What you lose leaving Plausible

• Bot filtering (don't count crawlers as visitors) needs an up-to-date bot signature list
• Real-time visitor counts at genuinely high traffic need a bit of care about write contention on the counter
• Their funnel and goal-tracking features add real query complexity beyond raw pageviews

What you lose leaving Mixpanel

• Funnel analysis (what % of users completed steps A then B then C, in order) is a genuinely non-trivial query pattern at scale
• Retention cohort analysis (of users who did X in week 1, how many came back in week 4) compounds that complexity
• Keeping ad-hoc queries fast over billions of events needs a columnar or event-optimized database, not a general-purpose one

Questions