Why do dealership marketing numbers never agree between systems?

Name the authoritative source per measure

Units and gross come from the DMS. Opportunities and follow-up come from the CRM.

Impressions and clicks come from the platform. Reporting that mixes sources without saying so produces arguments, not insight.

Windows and models differ

Attribution windows, view-through counting and last-click versus multi-touch models all change the number legitimately. Ask each provider to state its window and model, then compare like with like.

Deduplicate before you conclude

The same customer arriving through several channels is counted several times. Without deduplication, the sum of channel-reported results exceeds the store's actual sales.

Reconcile a small set monthly

Pick three figures and reconcile them every month. Persistent unexplained variance is a data problem worth fixing; a new dashboard is not the fix.

A reconciliation exercise

Choose a month and list sales delivered from the DMS. For each, look in the CRM for its source, then in advertising and analytics for the touches recorded.

Note where the trail breaks: no CRM record, a wrong source code, a phone call not tracked. The exercise usually reveals two or three process gaps that account for most disagreement and can be fixed at the source.

Agreeing a reporting standard

Publish a one-page reporting standard that lists each measure, its source system, definition, window and owner. Require vendors to report in those terms alongside their own.

When numbers still differ, the standard tells you which is authoritative and why. It also helps new staff and vendors understand how the store measures itself.

Common mistakes in reconciling numbers

Adding platform-reported conversions to get a total, comparing periods with different windows, blaming a vendor before checking the tagging and rebuilding dashboards instead of fixing data are common. Another is expecting exact agreement. Set an acceptable variance, investigate outliers and record the explanation of any known differences.

Questions to ask each vendor

How do you define a conversion? What window and model do you use?

How do you deduplicate across channels? How do you match to sales?

What do you not measure? Ask for the answers in writing and compare them to the store's reporting standard so differences can be explained rather than argued.