Signal-to-CRM Attribution Reconciliation (Retention)

Signal-to-CRM Attribution Reconciliation (Retention)

1. 01

   Freeze the intent source list and record the exact signal taxonomy in use for the account

   Log which signal types feed the pipeline (topic surge, competitor research, community post, job posting, paid-demand task) and the vendor each comes from, so a later discrepancy can be traced to one source rather than the whole stack.

2. 02

   Pull the raw signal export for the reconciliation window and stamp every row with its source and ingest timestamp

   A 30-day window works for most retainers. Rows without an ingest timestamp cannot be reconciled and should be quarantined before they reach the CRM.

3. 03

   Match each signal to a CRM account using a deterministic key, not a fuzzy company-name join

   Domain plus a secondary identifier (LinkedIn URL or account ID) cuts false matches. Name-only joins on companies like 'Vector' or 'Clearbit' style generic names produce duplicate account records.

4. 04

   Classify every matched signal as either corroborated or orphaned against the client's actual deal stages

   Corroborated means the account shows a matching stage movement (discovery, demo booked, proposal) within 45 days of the signal. Orphaned means the signal fired but no pipeline movement followed.

5. 05

   Compute the orphan rate by source and flag any source above the agreed noise ceiling

   A source that produces more than 60 percent orphaned signals in a month is contributing volume, not intent. Set the ceiling with the client in writing before the first reconciliation.

6. 06

   Write the reconciliation summary into the client-facing report with source-level orphan rates and one recommended action per source

   Keep it to one page. The client needs to see which sources earned their place in the retainer and which are being cut, not a dashboard export.

7. 07

   Apply the source changes to the live configuration and schedule the next reconciliation before closing the ticket

   Removing a noisy source without scheduling the follow-up check means the same source creeps back in during the next campaign build.