The practical takeawayReconcile a small sample of individual records after aligning the event, date basis, and counting rules.
- Compare definitions
- Match the records
- Check the timing
Your advertising dashboard shows one number of conversions. Your CRM shows a different number of leads. The immediate temptation is to decide which tool is wrong. A more useful first question is whether the reports are measuring the same event over the same period with the same rules.
A conversion might represent a form submission, a call, or a later qualified opportunity. A CRM might count unique people, newly created records, or deals. Those totals can differ even when data is flowing correctly. They can also expose genuine problems, including missing inquiries and duplicate tracking.
The goal of reconciliation is to understand the difference well enough to make decisions, then fix the gaps that represent missing or incorrect information.
Define exactly what you are comparing
Write a short comparison statement before opening more reports. For example: “Unique website quote inquiries created during October, including all marketing channels, excluding test and spam records.” Then check whether each report can actually represent that definition.
List the conversion actions included in the ad report. A total that combines quote requests, brochure downloads, and telephone clicks cannot be compared directly with qualified CRM opportunities. Break the total into its component actions before trying to explain the gap.
Also check whether the CRM report counts contacts or submissions. One person can submit two inquiries, and one inquiry can create multiple related records. Neither structure is automatically wrong, but the counting unit must be explicit. Keep a small reporting dictionary so future meetings do not restart this discussion.
Align the calendar and date basis
Two reports can use the same visible date range and still assign events to different days. A system may organize results by the advertising interaction, the inquiry, the qualification event, or the time a record was imported.
Google Ads documents that its standard conversion reporting can assign a conversion to the earlier ad click, while conversion-time columns organize it by when the action occurred. Its offline conversion troubleshooting guidance also explains why upload dates and account time zones require care when comparing reports.
For an illustrative example, a person might click an ad on Monday, inquire on Wednesday, and be marked qualified the following week. Those are three legitimate dates. Choose the one that answers your business question, then label the report accordingly.
Allow time for the business process and integration to catch up. A recent campaign cohort may still have unresolved inquiries. Comparing its final sales rate with a much older cohort can create a misleading picture of performance.
Inspect attribution and counting rules
An advertising platform asks which outcomes it can attribute to advertising under its own rules. Your CRM may describe the original known source, the most recent touchpoint, or a manually chosen source. Attribution windows, models, and available identifiers can change the resulting totals.
Record which rules each report uses rather than assuming their labels are equivalent. A field named “source” is not a complete definition. Ask when it is populated, whether it can be overwritten, and what happens when several interactions are available.
Remember that clicks and visits are also different measures. Google describes several reasons for differences between ad clicks and Analytics sessions in its clicks and sessions troubleshooting guide. Do not use equality between those totals as the only test of whether tracking works.
Agree on what each system is trusted to answer. The CRM can describe recorded lead status and sales activity. Website analytics can describe measured onsite behavior. Advertising reports can support platform-specific campaign analysis. A shared dashboard should preserve those definitions.
Trace a small sample of individual records
Once definitions are aligned, inspect a manageable sample of records with known paths. Include an ordinary inquiry, a repeated inquiry, a call, a test record, and an inquiry with missing source information. Use internal identifiers and restrict access to the people who need to investigate.
- Did the website accept the inquiry once?
- Did the receiving system create the expected record?
- Were the available source fields retained?
- Did the record meet the qualification rule?
- Did an integration send the intended outcome, and did the destination accept it?
Look at integration delivery and error logs rather than relying only on a dashboard total. A successful connection setup does not prove that every subsequent record arrived. Document the failing example, the expected result, and the owner of the fix.
Make reconciliation part of ongoing care
Create a short exception list: duplicates, missing source fields, rejected updates, and records waiting for a staff decision. These categories help the team separate technical fixes from process changes. Track whether exceptions are increasing after a website, CRM, or campaign update.
Keep the reporting definitions and integration map beside the dashboard. When someone changes a form or qualification stage, review the impact on the measurement chain. This is easier to maintain as an ongoing responsibility than as a recurring emergency before a board meeting.
Bring TripleSkinny a specific comparison, the relevant date range, and a few safely shared examples. That gives the investigation a clear starting point: explain the known differences, repair the broken handoffs, and make the reports useful for the decisions you need to make.



