Why Google Ads, GA4 and your CRM show different numbers

Use a diagnostic table before changing tracking

These are starting hypotheses, not verdicts. The Google Ads discrepancy guidance describes several legitimate reasons for differences. Your account settings and integration behavior determine which ones apply.

Write down what each number counts

A reporting discrepancy is a difference between measures that appear comparable. Start by identifying their unit: ad-platform conversion, analytics event, unique inquiry, qualified opportunity or won deal. A person submitting a form twice can produce multiple events but one CRM inquiry. That is not necessarily a failed integration.

Create a small reconciliation sheet with the report name, event definition, date basis, timezone, attribution settings and counting rule. Preserve the original totals. Do not change conversion settings simply to make a screenshot match another dashboard.

Reconcile a small sample before the whole dashboard

Choose a handful of inquiries for which your team can inspect the full path. Include a normal submission, a repeat submission and an inquiry that became a customer later. For each, record the event name, timestamp, available identifiers and the business outcome. Keep the original record from each system beside your explanation.

Suppose a customer submits twice but creates one sales opportunity. Two form events and one opportunity can both be correct. If a tag fires twice for one submission, however, that is an implementation defect. The totals alone cannot distinguish these situations; the individual records can.

When the sample is understood, turn each explanation into a repeatable check. Group known differences such as counting units and reporting dates separately from missing events or duplicates. Keep unresolved differences visible with an owner. Do not change conversion definitions merely to make two dashboards agree, because the new agreement can conceal a less useful measurement.

Keep a reconciliation worksheet with these columns

• Source system, event or stage definition, and the unit being counted.
• Event date, reporting timezone and the comparison period.
• Available inquiry or event identifier and the related CRM opportunity.
• Expected difference, confirmed defect or unresolved question, with the next check.

Compare the same group of inquiries

Suppose an illustrative October report contains inquiries created in October and deals won in October. Those deals may include September inquiries, while some October inquiries will close in November. Dividing October wins by October inquiries mixes populations.

For a lead-cohort report, select inquiries created in a defined period and follow their later outcomes. For a cash report, use actual payment dates. Both can be useful, but they answer different questions. Record the observation date so an open cohort does not look final.

Trace examples across the system

1. Choose a small set of permitted records representing a match, a missing record and a duplicate.
2. Follow source context from the initial interaction through form or call intake.
3. Check the stable inquiry identifier and the CRM stage history.
4. Compare the event timestamp and the report’s date basis.
5. Record the explanation or mark the case unresolved with an owner.

Inspect the smallest amount of data needed. Do not put personal contact fields into analytics events or public debugging screenshots. A useful investigation can usually work with internal record references and a documented event path.

Decide what needs fixing

An expected difference needs a reporting explanation. A missing CRM record needs an intake repair. A lost identifier needs a source-capture repair. A duplicate needs a counting and event-identity rule. Keeping those actions distinct avoids rebuilding the entire measurement setup for a definition mismatch.

The acceptance test is not that every total becomes identical. It is that the team understands the definitions, confirmed delivery failures are addressed and the remaining unknowns stay visible. Growth Control implements that connection between source evidence and the operating sales process.

Questions and answers