When Google Ads and Your CRM Disagree, Check the Dates First

Google Ads and the CRM may be using two different clocks

Standard Google Ads conversion reporting generally organizes an attributed conversion by the time of the ad impression or click. Google also provides columns such as Conversions (by conv. time)for reviewing results by the time the conversion itself happened. The CRM usually groups a record by a business event such as lead creation, qualification, appointment, opportunity close, or payment.

Google's own guidance for offline conversion discrepancies tells advertisers to compare the reporting date, date range, and time zone, and not to compare by upload date. See the official guidance on offline conversion import discrepancies.

Neither date is universally better. Interaction time helps explain which campaign created the opportunity. Conversion or close time helps explain what the business completed during the period. A useful review labels the clock instead of blending the two.

Build one comparable cohort before comparing totals

Choose a group of advertising interactions or inquiries and follow those same records forward. Do not compare every conversion shown by Google in July with every sale closed in the CRM in July unless the reporting rules confirm that those are the same customers.

1. Choose the starting event. Use an ad interaction date, website inquiry date, or another precisely defined event.
2. Allow for the sales cycle. Keep the outcome window open long enough for the selected inquiries to qualify, book, close, or become lost.
3. Match individual records. Use an appropriate click identifier, lead identifier, email, phone number, or another permitted first-party identifier rather than matching totals alone.
4. Record the final known status. Distinguish no response, unqualified, estimate issued, booked, lost, completed, and paid when those stages matter to the decision.

A cohort view answers a better question: "What eventually happened to the inquiries created during this period?" A close-date view answers a different one: "What did the business close during this period, regardless of when the demand began?" Both are useful when they remain separate.

Check timing, identity, definition, and duplication

A date adjustment cannot repair every disagreement. Validate four dimensions together.

Timing

Compare the interaction date, conversion date, lead-created date, close date, attribution window, sales cycle, account time zone, and any import or processing delay.

Identity

Confirm that the platform event and CRM record can be linked to the same person or opportunity using an appropriate identifier. Similar dates and equal totals are not identity proof.

Definition

Write down what Google counts as a conversion and what the CRM counts as a lead, qualified opportunity, sale, and collected payment. Similar labels may describe different events.

Duplication

Check repeat form submissions, multiple tracked actions from one person, repeated calls, merged CRM records, and import rules. Google also recommends a transaction or order ID to help control duplicate offline conversions.

Google documents how offline outcomes can be matched back to advertising interactions and recommends qualified- or converted-lead goals for deeper funnel events. The implementation should follow the business's consent, privacy, and data-governance requirements. See Google's guidance on qualified and converted leads.

Let the pattern determine the next investigation

Once the records are comparable, different disagreements point to different work.

Same people, different months

The primary issue is reporting timing. Preserve both interaction-date and outcome-date views, then use the one appropriate to the decision.

Google event, no CRM record

Test the tag, form, call threshold, routing, spam handling, and record creation. Determine whether an event fired without a usable inquiry or a valid inquiry failed to reach the CRM.

CRM record, no ad match

Inspect identifier capture, consent state, attribution rules, imports, unsupported or expired identifiers, and other acquisition sources. Do not assign every unmatched lead to paid marketing.

Records match, outcomes are weak

The measurement may be working. Review search intent, targeting, offer, landing experience, lead quality, response time, sales process, pricing, and capacity based on where the matched records stop.

When the dates are not the problem

Do not let a timing explanation become a permanent reason to avoid a performance decision. Dates are not the main problem when individual records have been matched, the same event definitions and cohort rules are in use, the sales cycle has had time to mature, and the business outcome is still weak.

Other evidence may be more important when:

• Search terms, audiences, geography, or ad promises do not fit the intended customer.
• The landing page creates a different expectation from the advertisement.
• Tracking fires without a meaningful customer action.
• Relevant inquiries are not answered, qualified, or followed through consistently.
• The CRM cannot preserve source, identity, or status after the initial inquiry.
• The campaign economics remain unacceptable after valid outcomes are reconciled.

A campaign problem and an internal record problem can exist at the same time. Finding one does not excuse the other.

Use one shared review instead of two competing reports

A dependable monthly review can show both views: campaign activity grouped by interaction date and mature business outcomes grouped by their actual outcome date. It should also identify how many records matched, which did not, what definitions were used, and which recent cohorts are still developing.

When the discrepancy belongs in campaign configuration or management, that is a Google Search Ads management question. When source, identity, status, and outcome cannot survive the internal path, the evidence problem may belong in business systems consulting. The records should meet for diagnosis without turning those into the same service.