AfieldIn this article
Before importing prospects into your CRM, inspect a small sample. A repeated domain may identify a duplicate company, but two people at that company can still be distinct contacts. Keep the original file so cleanup decisions can be checked later.
This is a spreadsheet-preparation procedure. Afield’s current capability set supports company and public-contact research; this guide does not promise a bulk CRM import or automatic record merge.
Inspect four rows before the full file
Fictional sample. The example domains and contacts below are not real prospects.
companyName,domain,contactName,title,email,country
Demo Packaging,demo-packaging.example,Alex,Engineering,,Germany
Demo Packaging GmbH,demo-packaging.example,Alex,Engineering,,Germany
Demo Packaging,demo-packaging.example,Robin,Purchasing,,Germany
Demo Components,,Casey,Sales,,Germany| Rows | Decision | Reason |
|---|---|---|
| 1 and 2 | Review as a possible duplicate. | The domain, person and role match. Confirm the legal entity before keeping one row. |
| 1 and 3 | Keep the two contacts separate. | The company matches, but the people and roles differ. |
| 4 | Hold outside the import. | The company identity is unresolved without a domain or another reliable identifier. |
After review, this sample may contain one company with two contacts and one unresolved row. Record that expected result before importing. Otherwise a CRM creating four accounts can look superficially successful.
Research the missing identity, not the entire list again
Use a narrow Afield request for the unresolved company, supplying any location, product or source you already have:
Research this company identity: [name, location and known source].
Find its official website and explain the match. For [buying role],
find a current public contact route. Keep uncertain matches separate.
Do not infer an email address from a naming pattern.Transfer only checked facts back to your preparation sheet. A public role is not the same as an available email, and a shared company inbox is not a named person’s address. Follow the contact-gap procedure when a row remains incomplete.
Keep history outside fields that cannot represent it
Retain source URL, observation date and cleanup decision alongside the original file. Previous replies, contacted status and account notes may require different CRM fields or a separate migration procedure. Do not treat a CSV row as a substitute for its underlying conversation.
In the destination CRM, map each source column to a supported field and use its documented duplicate behavior. Check whether a company identifier, contact email or another key determines updates. If the intended behavior is unclear, stop at preview rather than relying on an automatic merge.
Verify the sample in the destination
- Import the reviewed sample using your CRM’s own supported workflow.
- Open the resulting company and both contact records.
- Compare names, domain, roles and blank fields with the prepared file.
- Check that neither contact was collapsed into the other and that the unresolved company was not added.
- Continue with another batch only after this result matches the expected structure.
If the destination has already created a record, inspect it before retrying the whole batch. For complex organizations, read parent and subsidiary research. For a new market with no existing file, begin with company discovery.