Why ad conversions and sales records differ

01

Inventory the conversion events

List each event included in the ad report: form submission, call, appointment or sale. Check whether it counts one or every conversion after an interaction, its attribution settings and conversion window. Google notes that its conversion column can include modeled conversions; that total may not match identifiable records one for one.

In the sales system, define a distinct inquiry and a confirmed sale. Record the date each report uses and whether refunds, duplicates or invalid contacts are excluded. A form submission is an inquiry event, not proof of a customer.

02

Reconcile a closed cohort

Choose a period old enough for the sales cycle and conversion reporting delay. Match records using permitted identifiers, then group unmatched events by likely cause: duplicate submissions, missing identifiers, offline sales not recorded, different dates or attribution. Do not force a match where the evidence is absent.

Hypothetically, an ad report shows 100 form events while the sales system has 90 distinct inquiry records linked to the campaign. Investigate the difference of 10. Do not assume all 10 are duplicates or lost inquiries. If 9 of the 90 became customers within the agreed window, the observed sale rate for those matched inquiries is 10%, with unmatched events reported separately.

03

Repair the reporting path

Fix confirmed errors at their source, such as a form firing twice or a missing campaign field. If importing later sales outcomes, follow the platform's current requirements and applicable privacy rules. Document which outcomes were uploaded and when.

Keep a reconciliation log and review it after a change. Attribution assigns reporting credit under a model. It does not establish that an ad caused every credited sale. Use consistent definitions and closed outcome windows when comparing campaigns.

04

Work through an exception

Suppose a person submits a form twice after one ad interaction and the sales team merges both submissions into one inquiry. The platform may count one or two events depending on its settings, while the sales system reports one record. Check the event identifiers and counting settings before deciding whether a tracking repair is needed.

Also inspect sales without a known campaign and campaign events without a matching record. Some gaps cannot be resolved because identifiers were not available or modeled conversions are included. Show an unexplained remainder rather than assigning it to a campaign or deleting it from the report.

05

Match date and attribution before matching records

A Google Ads report can attribute a conversion to the earlier interaction date, while a sales record uses the date the inquiry or sale occurred. Google also provides conversion-time columns. Compare the same event using both date views before diagnosing a gap. Neither view makes platform credit identical to proof that advertising caused a sale.