Mixpanel

Mixpanel, in 30 seconds

Mixpanel is the polished product analytics platform that lets non-technical product-led growth teams answer behavioral analytics questions about funnels, retention, and cohort analysis without writing SQL. In May 2026, the Free plan covers 1 million events monthly with 10K session replays, Growth charges $0.28 per 1,000 events past that, and Spark AI now ships natural-language reports across Boards, Reports, and Funnels. It rates 4.5/5 on G2.

• Freeup to 1M events/mo
• 4.5/5G2 score, 1.2k+ reviews
• ~8Kpaying customers
• 2009founded by Suhail Doshi

How much does Mixpanel cost per event?

Mixpanel switched fully to event-based billing in early 2026, replacing the old MTU pricing model that confused most buyers. Here is what each tier actually buys you, pulled from Mixpanel's pricing page.

The math is friendlier than it looks at small scale and brutal at large scale. At 5M monthly events you owe roughly $1,120/month on Growth. At 10M events, that climbs to about $2,520/month. Past 50M events you should phone enterprise sales because per-event rates negotiate down sharply on annual commits.

Where the bill creeps: data pipelines, feature flags, extra session replays beyond included caps, and Group Analytics for B2B account-level reporting are priced as add-ons on Growth. Many teams budget for a base seat, then watch event volume balloon as they add server-side tracking. Track only the events you will actually query.

Is Mixpanel right for product-led teams?

The sweet spot is product-led SaaS between 5 and 500 employees where product managers, growth marketers, and designers all need to self-serve behavioral analytics without a data team in the loop. If you ship features weekly, run onboarding experiments, and care about activation-to-retention curves, Mixpanel is built for your daily workflow.

It is also a fit for mobile-first consumer apps. The native iOS and Android SDKs are mature, user journey reports surface drop-offs cleanly, and the cohort builder lets growth teams slice retention by acquisition channel or feature exposure. Companies like Uber, Yelp, BuzzFeed, and Lemonade run on it.

It is the wrong tool for engineering teams who want self-hosting, marketing teams who need full ad-spend attribution, or massive enterprises that already pay for warehouse-native analytics. If your event volume sits above 100M per month and you already centralize data in Snowflake, the math will push you toward warehouse-first competitors.

What does Spark AI actually deliver?

Spark AI, Mixpanel's generative AI layer (originally launched as Mixpanel AI in 2024 and rebranded to Spark in 2025), is now generally available across Insights, Funnels, Retention, and Flows. You ask in plain English — "show me weekly active users in the past 90 days broken down by signup country" — and Spark builds the report, picks the chart, and lets you edit it like any saved board.

The 2026 additions are the parts worth caring about. Spark now supports anomaly detection on saved reports, so you get a Slack ping when conversion drops 15% week-over-week without setting alert thresholds by hand. AI-powered session replay summaries condense a 12-minute replay into three bullet points. Magic Playlists group replays by criteria like "rage-clicked the upgrade button."

What is still beta in May 2026: agentic root-cause analysis (Spark proposing why a metric moved, not just that it moved) and the experimental SQL-export bridge for warehouse teams. Treat anything labeled beta as a preview, not a production workflow.

Mixpanel vs Amplitude vs PostHog vs Heap

Watch Mixpanel Spark in action

The walkthrough shows former Mixpanel CEO Amir Movafaghi typing a question and watching Spark build a funnel report in under five seconds. It is a fair representation of the daily Spark workflow today, though the 2026 UI now shows confidence scores beside each suggested chart.

Where Mixpanel disappoints

Pricing creep is the dominant Reddit and G2 complaint in 2026. Buyers say the jump from Free to Growth feels steep, and once a team adds server-side events, mobile SDK auto-events, and revenue tracking, monthly bills routinely outrun the original budget by 2-3x. A common refrain: "they get you in cheap, then events build up and your $20 month becomes $150."

The second weakness is the learning curve. Casual users — folks who just want a weekly active user count — find the Insights builder over-engineered for their needs. Compared to Amplitude's newer onboarding, Mixpanel feels more demanding for the first two weeks. G2 reviewers consistently note the steep curve on advanced funnels and Metric Trees.

Add-on creep is the third. Group Analytics for B2B reporting is essentially required for product-led B2B SaaS but priced separately. Data pipelines to ship events into a warehouse cost extra. Several teams told us they cut tracked events deliberately to keep the bill flat — which defeats the point of buying behavioral analytics.

The decision matrix

Your next move