Prepare your next client review with a clearer picture of tracked activity. Connect PostHog to Catchr, then use ChatGPT or Claude through Catchr MCP to explore event counts, campaign tags, page URLs and device patterns.
For freelancers and agencies, it is a practical way to investigate changes and build talking points from the events a client already tracks.
“Compare paid media performance across my client accounts and flag the biggest changes.”
PostHog
Google Ads
GA4
Used by the world's leading companies
Start with an event name and reporting period, then break down event counts by time, UTM tags, page, geography or device. Records measures the number of events, so keep conclusions tied to tracked activity.
Use Event Name and Records to see which tracked actions occur most often. Group counts by date, week, month or hour to examine changes before a client meeting.
Records represents one event in detailed results and the event count in grouped results. Sum it to report total events; it does not represent unique people.
Break down a selected event by UTM Source, UTM Medium and UTM Campaign to compare activity associated with each tag. Add Current Url to see which pages appear in those events.
Use these breakdowns to discuss campaign-tagged activity with clients. UTM values describe tags attached to events; they do not establish which campaign caused an action.
Compare event counts by Device Type and Browser to identify differences worth investigating. Country Code and Region add geographic context to the same tracked activity.
Device Id is also available for examining events associated with a particular device identifier. Keep device-level findings separate from claims about individual people.
A focused event analysis, a recurring client report and a question through Catchr MCP serve different needs. Choose the approach that fits the PostHog question you need to answer.
Use a report for a consistent view of PostHog activity and Catchr MCP for follow-up questions, such as which URLs contributed to an event-count change or how campaign-tagged events differ by device.
Connect PostHog in Catchr using an API key and the base URL for your PostHog instance. The URL must match your US, EU or self-hosted deployment, and scalar reporting requires the query:read permission.
Select the relevant connected account and connect your AI destination through Catchr MCP. Start with a specific event name, date range and breakdown so your first question has a clear reporting scope.
Replace the bracketed values with your client account, dates and reporting choices. Each prompt asks for a defined event breakdown you can review before using it in a client discussion.
Using Catchr, compare summed Records by Event Name for PostHog account [account] between [baseline start date]–[baseline end date] and [comparison start date]–[comparison end date]. Show each period’s event count, absolute change and percentage change. Flag changes above [percentage threshold], and mark percentage change as undefined when the baseline is zero. Describe these as event counts, not unique users.
Copy promptUsing Catchr, for PostHog account [account] from [start date] to [end date], filter Event Name to [event name] and sum Records by UTM Source, UTM Medium and UTM Campaign. Rank the combinations by event count, show missing tags separately, and list combinations with at least [minimum event count] events. Summarize which tags have the most recorded activity without making causal attribution claims.
Copy promptUsing Catchr, for PostHog account [account], filter Event Name to [event name] and compare summed Records by Current Url between [baseline start date]–[baseline end date] and [comparison start date]–[comparison end date]. Rank URLs by absolute event-count change, show increases and decreases, and keep missing URLs separate. Identify the pages contributing most to the overall change.
Copy promptUsing Catchr, analyze [event name] in PostHog account [account] from [start date] to [end date] for Country Code [country code]. Sum Records by Device Type and Browser, show each combination’s share of the selected event total, and flag combinations with at least [minimum event count] events. Keep missing device or browser values visible and suggest follow-up questions without treating differences in event volume as evidence of a technical fault.
Copy promptRecords measures events. A detailed row represents one event; a grouped row contains its event count. Sum Records for total events, and do not label that total as unique users, sessions or customers.
You need an API key and the explicit base URL of your PostHog instance, whether it is a US, EU or self-hosted deployment. Scalar reporting requires query:read permission. Then select your connected account and use ChatGPT or Claude through Catchr MCP.
Yes. Filter by Event Name and compare event counts using UTM Source, UTM Medium, UTM Campaign and Current Url. These values describe what was attached to the events; missing tags limit the breakdown, and the results do not establish causal attribution.
You can compare counts for a selected event by Device Type and Browser, with Country Code or Region for geographic context. Differences can guide further investigation, but event volume alone does not establish a browser fault or explain why activity changed.
Connect your sources to Catchr, add the MCP server to your AI assistant, and ask about the work that matters now.
