Explain missing rows, withheld data, and "why numbers don't match"
GA4 can apply thresholding to protect user privacy and sampling when report/exploration volume exceeds limits. This recipe produces a plain-English explanation, detection checklist, and mitigation playbook (including how to change the question—not just the dashboard).
Diagnose whether we're seeing GA4 thresholding or sampling, then propose mitigations. Output: - Diagnosis with reasoning (thresholding vs sampling vs something else) - Mitigation plan (what to change in queries/reports) - One-paragraph explanation for non-technical stakeholders - A reusable "data limitations" note for recurring reports Inputs: - Report/exploration URL or description: - Dimensions/segments used: - Date range: - What looks wrong (missing rows, low numbers, warnings):
This recipe helps teams stop "debugging ghosts" when data is withheld or sampled by design.
Diagnose missing conversions and "flying blind" measurement fast
Use this when the numbers don't match: ad platforms over/under-report, GA4 looks off, CRM revenue doesn't reconcile, or privacy changes (ATT/cookie loss/consent) have degraded tracking. It produces a root-cause shortlist, a "what to trust" guidance note, and a prioritized fix plan.
Stop reacting to incomplete data from the last 24–48 hours
GA4 reporting can be delayed, and some reports can be incomplete due to processing latency. This recipe sets "freshness rules," creates a monitoring checklist, and defines when to use real-time vs standard reports vs backend truth.
Label mistakes so patterns become obvious
Traders often report that profitability improved only after tracking mistakes (not just P&L). This recipe forces a mistake tag on every trade and compiles a mistake leaderboard.
Converts tags + stats into one concrete rule change
Traders often recommend a weekly review to spot repetitive patterns (revenge trades after first loss, overtrading during lunch, etc.). This recipe compiles the week into a short brief and proposes one fix.