wiki/knowledge/google-ads/ahs-conversion-tracking-audit.md · 655 words · 2025-11-19
AHS Conversion Tracking Audit
Overview
During the weekly sync on 2025-11-19, a critical data quality issue was identified in the Advanced Health & Safety (AHS) Google Ads account: conversion tracking is severely misconfigured, reporting thousands of conversions when the actual lead count is approximately 17–18. The root cause is that page views (and possibly other low-intent events) are being counted as conversion actions, inflating reported performance and likely corrupting any Smart Bidding strategies relying on that signal.
This issue was flagged by Mark Hope after a separate meeting with the AHS client the day prior. Gilbert Barrongo confirmed he was already pressing Anup to audit and clean up erroneous conversion actions across accounts.
See the source meeting notes: [1]
The Problem
| Metric |
Reported (Broken) |
Actual (Estimated) |
| Conversions |
Thousands |
~17–18 |
| Conversion Type Being Counted |
Page views + other non-lead events |
Form submissions / real leads |
Why This Matters
- Bidding corruption: If the account uses any automated or Smart Bidding strategy, it is optimizing toward page views rather than leads. This wastes budget and suppresses real lead volume.
- Reporting is meaningless: Any ROAS, CPA, or conversion rate figures shown in the account cannot be trusted until tracking is corrected.
- Client trust risk: If AHS is reviewing campaign performance dashboards, they are seeing wildly inflated numbers that do not reflect business outcomes.
Root Cause
Multiple conversion actions are active in the account, at least one of which fires on page load rather than on a meaningful user action (e.g., form submission, phone call, thank-you page). This is a configuration error — likely a Google Ads tag or Google Tag Manager trigger misconfigured to fire on all pages rather than a specific confirmation or lead event.
Action Items
- [ ] Anup — Audit all active conversion actions in the AHS Google Ads account. Identify and remove or pause any actions counting page views or other non-lead events. Confirm that only genuine lead events (form submissions, qualified phone calls, etc.) remain as primary conversion actions. (@Anup)
- [ ] Gilbert — Follow up with Anup directly to ensure this is prioritized and completed. (@Gilbert)
- [ ] Mark / Sebastian — After tracking is corrected, review historical performance data with the understanding that prior conversion figures are unreliable. Assess whether bidding strategies need to be reset. (@Mark @Sebastian)
Recommended Fix
- Audit conversion actions — In Google Ads → Tools → Conversions, list all active conversion actions and identify which events they correspond to.
- Pause non-lead actions — Any action tied to page views, session starts, or other engagement metrics should be set to "Secondary" or paused entirely so they do not influence bidding.
- Verify lead actions — Confirm that form submission or thank-you page triggers fire correctly and only once per genuine lead. Use Google Tag Manager's Preview mode or Google Ads' Conversion Status tool to validate.
- Reset Smart Bidding learning period — If automated bidding is in use, expect a learning period reset after conversion actions are corrected. Consider temporarily switching to manual CPC during this window to avoid erratic behavior.
- Establish a baseline — Once tracking is clean, document the true conversion volume (estimated ~17–18/month) as the new baseline for performance reporting.
Broader Pattern
This is not an isolated incident. Gilbert noted that multiple accounts have had conversion action configuration issues, and Anup has been tasked with a broader audit across the portfolio. This suggests a systemic gap in the conversion tracking QA process at account setup and during ongoing maintenance.
Related issue: [2] — Reynolds Transfer's "Madison Dane County" campaign also shows zero reported conversions despite $1,500+ in spend, which may similarly be a tracking misconfiguration (phone calls as conversions not properly captured).
References