Duplicate contacts and companies
Same person twice, or the same company under two names, each collecting activity after import.
CRM import
Clean the data before it becomes a CRM record
A successful import only proves the CRM accepted the file. Duplicate contacts, broken associations, and conflicting company names are harder to unwind after they are objects, automations, and reports — not rows in a spreadsheet.
Auditere is not a CRM and does not replace Salesforce, HubSpot, or your system of record.
The actual problem
Post-import cleanup is a different, slower job
Before import, a duplicate is two rows. After import, it can be two contacts, split activity, wrong ownership, and broken reporting. Investigating first is usually cheaper.
Most painful migrations are not “the importer failed.” They are files assembled from an old CRM export, a marketing list, and a spreadsheet someone has maintained for years — then mapped in a hurry.
Vendor tools can validate format, required properties, and unique identifiers. They are not a substitute for deciding whether two similar people are the same customer.
What can go wrong
Issues that become CRM objects
Missing or damaged identifiers
Blank emails, reused IDs, Record IDs from the wrong account, leading zeros dropped in Excel.
Mapping that is textual, not semantic
A column called Company that means legal name in one file and parent account in another.
Inconsistent phones, emails, and countries
Valid-looking values that still fail contactability or association rules.
Blank fields the business actually needs
The CRM may allow a sparse contact. Your process may not.
Stale and conflicting sources
Three versions of the same customer, each “true” in a different export.
Example scenario
Three sources, one customer, no safe auto-merge
A team combines a previous CRM export, a billing CSV, and a conference spreadsheet before moving into a new CRM.
| Source | Name | Company | |
|---|---|---|---|
| Old CRM | Sam Ortiz | [email protected] | Nimbus Analytics |
| Billing | Samantha Ortiz | [email protected] | Nimbus Analytics Ltd |
| Event list | S. Ortiz | [email protected] | Nimbus |
Automatically keeping the newest row might drop the work email. Automatically merging on last name might attach a personal inbox to a company record. The investigation is which fields are current, which are historical, and which associations must survive.
What to investigate
Identity, mapping, then relationships
What is a row?
Lead, contact, company, deal, or a mix? Importing the wrong object type creates cleanup that no dedupe rule will fix cleanly.
Which keys are trustworthy?
Email, domain, external ID, and vendor record IDs each fail in different ways. Confirm uniqueness and grain before upsert.
What must stay distinct?
Two contacts at one company can be correct. Two companies that share a domain can be correct. Similarity is not a merge key by itself.
What will the CRM accept anyway?
Import validity is not data validity. Outliers, stale owners, and contradictory statuses often load without errors.
Where Auditere fits
Investigate the source while it is still rows
Use Auditere to surface duplicates, missing values, suspicious records, and conflicting fields, then decide what should proceed into the CRM.
The workflow remains detect → review → resolve → verify. That is slower than “upload and hope,” and usually cheaper than unpicking production CRM records.
Deeper vendor-specific notes: checks before a Salesforce CSV import and preparing a CSV for HubSpot. The CRM-agnostic list is how to clean a CSV before CRM import.
Review the file before the CRM has to live with it
Start a Pro trial and investigate structured import files with explicit review — not silent auto-merge.