CRM Operations
Move Leads from Excel to a CRM Without Losing Notes or Creating Duplicates
The safest move starts with a small sample and a clear field map—not a full spreadsheet upload on Friday evening.
The spreadsheet has worked for years. Then a second salesperson joins, another tab appears, and “final leads latest” becomes one of four files. Moving to a CRM can help, but uploading the file exactly as it stands often carries the confusion into a new system.
A careful Excel to CRM migration has three jobs: preserve useful history, give every field a clear meaning and prove that the imported records match the source. You do not need a huge migration project. You do need a controlled first batch.
Keep an untouched source copy
Save the original workbook before cleaning it. Record which file, sheet and date the migration uses. Work on a separate copy and nominate one person to own changes while the import is in progress.
Agree how the team will handle new enquiries during the move. Otherwise, a salesperson may keep editing the old workbook while another starts updating the CRM, leaving two competing versions of the same customer.
Build a field map before touching the upload button
| Spreadsheet column | Destination meaning | Preparation |
|---|---|---|
| Mobile | Primary contact number | Store as text; retain country code. |
| Requirement | Current buyer need | Keep detail rather than reducing everything to a tag. |
| Status | Agreed pipeline stage | Translate inconsistent labels into defined stages. |
| Follow-up | Next action date | Use an unambiguous date format and timezone. |
| Remarks | Notes or supported custom field | Preserve meaningful history and line breaks. |
| Assigned to | Existing CRM user | Resolve spelling differences before importing. |
Not every destination supports every source field. Check the actual import mapping screen. If a column cannot be mapped, retain it in the source archive and agree a supported destination before proceeding. Never silently drop it.
Clean identity carefully
Phone numbers are identifiers, not quantities. Spreadsheet software can remove leading zeros or display long numbers in scientific notation. Format the column as text and inspect the stored values, not only their appearance.
Normalize formatting consistently, but do not guess missing country codes for an international list. Review duplicates using more than a similar name. Two employees may share a company switchboard; one buyer may have several enquiries. Decide whether you are importing people, enquiries or both.
Keep the original source and original creation date where supported. A migration date should not accidentally make old enquiries look like fresh marketing leads.
Use a deliberately awkward trial batch
Select a small sample that includes a normal record, an international number, a repeated contact, a blank optional field, a long note and a date. A sample of only perfect rows cannot reveal the failures that matter.
- Upload the sample using the supported file format.
- Review field mapping and destination pipeline.
- Inspect any validation or duplicate warnings.
- Import only after the review matches your intended result.
- Open the resulting records and compare values with the source.
Confirm what happens on a repeated upload before rerunning a batch. Some importers update existing records; others create new ones or reject them. Use the behavior of your actual configuration, not an assumption.
Reconcile before retiring the spreadsheet
Record submitted, created, updated, skipped and failed rows. These outcomes should account for the source batch. Investigate missing rows and check notes, owners and dates on a sample after the full import. Keep the original file read-only until the team accepts the result.
Once the data is reliable, agree on pipeline stages and exit rules. A clean import cannot fix a process where nobody knows what “warm” means.
Plan your first batch with SHVYA
Book an Excel migration discussion. Bring a redacted header row and a few fictional sample records. We can discuss the field map, ownership and validation steps before your real customer data moves.
