PostHog Ships Survey Translations, Markdown Notebooks and Faster Error Tracking.

PostHog announced on LinkedIn that its iOS and Android SDKs now support survey translations, allowing product teams to collect multilingual user feedback through native mobile apps. The update expands on PostHog's existing survey capabilities by letting developers serve localized survey questions based on device language settings.

The timing aligns with PostHog's broader push into integrated product development tools, following recent integrations with full-stack platforms.

PostHog's collaborative notebooks feature now renders Markdown pasted directly into the editor. Product teams can copy formatted text from GitHub issues, Notion documents, or technical specs and see it immediately rendered with proper headings, lists, and code blocks. The change removes a formatting step for teams documenting experiments, analyzing feature rollouts, or sharing insights across distributed teams.

The platform also added cohort edit history and activity logs, giving teams visibility into how user segments evolve over time. Teams can now audit changes to cohort definitions, useful when debugging experiments or tracking down why a particular segment's behavior changed.

Error Tracking queries now run significantly faster while consuming less memory, according to the update. PostHog did not specify exact performance gains, but the improvement addresses a common bottleneck for teams processing large volumes of error data. The optimization should reduce query timeout issues for organizations tracking errors across multiple services.

The platform added pganalyze as a signal source, letting teams correlate product analytics with PostgreSQL database performance metrics. Teams running PostHog alongside pganalyze can now surface database slowdowns or query patterns that coincide with user behavior changes or feature rollouts.

A new batch export API for file downloads allows teams to programmatically extract analytics data in bulk. The API complements PostHog's existing data warehouse integrations for teams building custom reporting pipelines or feeding analytics into business intelligence tools like Airtable or internal dashboards.

Developers can now test feature flag evaluation through Model Context Protocol (MCP), PostHog's framework for validating flag configurations before deploying to production. The feature helps catch misconfigured targeting rules or edge cases that could expose unfinished features to end users.

PostHog also made the default minimum detectable effect (MDE) for experiments configurable. Previously, teams used a platform-wide default when setting up A/B tests. Now product teams can set baseline sensitivity thresholds that match their statistical rigor requirements, whether detecting small conversion lifts or only tracking major behavioral shifts.

The updates ship as PostHog continues expanding beyond core analytics into workflow automation and developer tooling, including experimental AI agents for production monitoring. The platform remains open-source and self-hostable, targeting product teams that prioritize data ownership alongside analytics depth.