REVIEWS / ANALYTICS / OWNER INSIGHTS

🦉 WE READ 83 OWNER COMMENTS

Mixpanel: what owners actually say

Owners find Mixpanel useful for product analytics but the platform carries serious concerns around pricing transparency, data retention, and a recent security breach that cost them OpenAI as a customer.

HACKERNEWS · 68 REDDIT · 10 LEMMY · 3 PRODUCTHUNT · 2

What owners complain about

  • Opaque and punishing pricing SOME

    Multiple owners report the pricing model is confusing and even punitive: one user upgrading from the free tier was downgraded from 20M events/month to 10K events/month on a paid plan. Another dropped the product after the free startup package expired because paid tiers reached thousands of dollars per month.

  • Poor breach communication COMMON

    After a security incident, Mixpanel's own blog post was widely criticized as vague and defensive — omitting what systems were accessed, what data was exposed, timelines, and scale. OpenAI's disclosure to its own users was clearer than Mixpanel's direct communication. OpenAI subsequently dropped Mixpanel as a vendor.

  • Data retained after account closure FEW

    A user who closed their account months before the breach still received an incident notification, raising concerns that Mixpanel retains customer data well beyond account termination unless explicit deletion is requested.

  • Non-de-identified user data storage SOME

    Commenters flagged that Mixpanel stores data in non-de-identified form — device IDs are back-ported to internal customer records once a user is identified, creating a linked profile that raised privacy concerns in the context of the breach.

  • Adoption ceiling without dedicated analyst FEW

    One company adopted Mixpanel so PMs could self-serve product data, but realized they still needed to hire a dedicated analyst shortly after. Another org noted engineering sets up events but rarely revisits dashboards — real usage was PMs checking twice a week before standup.

What owners love

  • Individual-level user tracking

    Owners value Mixpanel for tracking user interactions at the individual level — something GA4 cannot do well since it focuses on aggregated, anonymous analytics.

  • Quick analytics for immature data stacks

    For organizations whose data architecture is still early, Mixpanel (alongside competitors like Amplitude) provides a fast way to get product and engagement analytics running without building a warehouse first.

  • Complex analysis hard to replicate in SQL

    Owners praise capabilities like Sankey diagrams showing common user paths from specific events, funnel analysis, and session-level insights that would be painful to reconstruct with raw SQL against a data warehouse.

  • Standard choice for European SaaS startups

    European startups and scale-ups commonly pick Mixpanel or Amplitude as their default product analytics tool, with product managers typically owning the setup and day-to-day usage.

Surprising patterns

  • Despite being positioned as a deeply embedded analytics platform, real-world usage at some companies is surprisingly light — PMs check dashboards maybe twice a week, usually just before standup when someone asks about feature adoption.
  • The breach highlighted that Mixpanel's core data model intentionally links anonymous device IDs to identified customer records, meaning a breach exposes far more than anonymous behavioral data — it exposes linked user profiles.
  • OpenAI's incident disclosure was significantly clearer and more detailed than Mixpanel's own blog post about the same breach, prompting commenters to note this as a textbook example of how not to handle vendor incident communication.

WHO SHOULD SKIP IT

Organizations with strict data-retention or privacy requirements — especially regulated industries — should think carefully, given the breach, data retention beyond account closure, and the platform's practice of storing identified user profiles rather than anonymized data.

2.0/10 GYIBB verdict
Full review →

Synthesised from 83 real owner comments across 4 platforms. Every point is grounded in the comments — no marketing, no AI guessing. How we do it →