Contact Data
Part of CRM data quality
Auditing data quality after a bulk import
Reconcile imported records, values, links and errors against the source population, then correct and recheck affected customer records.
Audit a bulk import by comparing the intended source population with what the CRM actually created, updated, linked and rejected. Do this while the source snapshot and import settings are available. A completion message does not show whether each customer record is usable.
Fix the audit boundary
Record the source file version, import time, object types, create or update mode, matching key and the person responsible for exceptions. Preserve the source population used for the run. If the file changed, establish which version the CRM received before investigating a difference.
Count source rows by intended outcome, not just file length. In a supported import mode, one contact may appear on two rows to establish two relationships while remaining one intended contact. Record expected new records, updates and associations separately, and rows held for review.
Reconcile outcomes in layers
| Layer | What to compare | What a mismatch may signal |
|---|---|---|
| Records | Intended creations and updates against CRM results | An unmatched key, unexpected new record or skipped row |
| Values | Important source values against destination fields | A rejected value, wrong mapping or overwrite |
| Relationships | Intended contact, account and deal links | A missing or incorrect association |
| Ownership | Intended assignee against current owner | An ambiguous user match or unintended transfer |
| Exceptions | Rejected or uncertain items against decisions | Work left without a reviewer |
HubSpot’s past-import view reports new records, updated records, new associations and errors separately. Its error guidance lets users review error types, affected columns and rows, and errored values. Inspect both the summary and affected records. Other CRMs need equivalent checks using the results they provide.
Pros and Cons of Using HubSpot’s Past-Import View
- ProsProvides clear breakdown of new records, updates, associations, and errors; includes error guidance with column and row details.
- ConsDoes not show individual record history before the import; requires manual verification for complex relationship chains.
Trace representative records
Select a new customer, an update to an existing customer, a contact linked to an account, a record with an owner and a known awkward case. For each, follow the source reference into the CRM and check the record ID, current values, links and open work. Include a blank optional value or a previously existing customer if those cases occurred in the import.
Inspect exceptions and groups that could conceal systematic errors: records with no owner, unexpected duplicate candidates, unmapped choice values and suspicious relationships. A random sample can find unanticipated problems; targeted checks address known failure modes. Neither proves every record correct.
Correct and recheck the affected population
Classify each issue before acting: source-data error, mapping error, uncertain identity, CRM validation error or wrong business rule. Give it an owner and a decision. Correct the cause, then re-import or edit only the affected records through a controlled route. Check the results again, including related work and later integration changes.
Treat deletion of records added by an import as separate from correcting records that existed before it; verify and correct any affected existing records directly. Keep the source snapshot and a correction log until the reconciled population and outstanding exceptions are clear.
Finish with a brief record of the population checked and comparisons made. Include unresolved exceptions and the person accepting the result.
