Turned PostHog off because its tracking made the app unusably slow
“it was collecting so much information that wasn't needed that it was making my app unusably slow”
2 people say they left PostHog; what else pushes users away, and what its category peers report
Friction holds 24 sourced reports about PostHog (posthog.com) from 77 public source threads (as of 2026-10-15); only 2 of them say the author left. This page shows those reports, what its users say pushes them away, and the problem profile of Umami, Matomo.
Get started free with your email Open PostHog on Friction
Where PostHog users say they went
Switching reports that name a destination. A count is the number of reports naming it, not a market share: Friction only holds what people say in public.
| Named alternative | Reports naming it | Newest report |
|---|---|---|
| A self-built or in-house tool | 1 | 2026 |
| self-hosted or EU-based alternatives | 1 | 2022 |
| Segment | 1 | 2021 |
The reports of leaving
Built own analytics because PostHog was overkill for simple attribution needs → self-built analytics
“PostHog was overkill for what I wanted (Country, Origin, UTMs, per-user attribution, entry page, revenue)”
What pushes people away from PostHog
The most-reported open problems, worst severity first. Each links to the item with its source.
S3 warehouse destinations fail with IAM-role authentication on Cloud US · Integrations · Blocks work
“Role-based AWS access is not available: BATCH_EXPORT_S3_EXTERNAL_ROLE_ARN is not set”
Training agent’s feature-query guidance encourages costly SQL · Performance · Blocks work
“On a large team, agents that start from it run out of memory”
Symbol-set uploads fail with Cloudflare R2 object storage · Self-hosting & upgrades · Blocks work
“error tracking symbol sets can never be uploaded”
Marketing site does not make clear what PostHog actually does · Setup & onboarding · Annoyance · 2 people
“I poked at the site for a few minutes, then had to ask an LLM what PostHog actually does”
Hobby self-hosted PostHog scales poorly and is hard to install · Self-hosting & upgrades · Annoyance
“The hobby offering doesn't have to scale poorly. It could scale just as well as your core offering”
Other product analytics tools on Friction, and what their users struggle with
Switching trades one set of problems for another. The same evidence, for each tool Friction covers in this category:
Umami
30 reports · 87 source threads · most-discussed area: Reliability & bugs
Matomo
19 reports · 69 source threads · most-discussed area: Reporting & analytics
Counts and reports reflect people who post in public (developers and technical buyers are over-represented). A tool with fewer reports is not a better tool; it may only be discussed less. How the data is built.