CRM data migration from spreadsheets or another CRM
Answer in brief
CRM data migration at an investment firm is mostly a decision exercise: what to keep, how fields map, which duplicates merge, how much history comes across, and when the old system goes read-only. Do the inventory and mapping first, load a sample, validate it with the people who know the relationships, then cut over on a fixed date. A CRM that fills itself from email, such as Canteris, reduces how much has to be migrated by hand.
Start with an inventory
Before mapping a single field, list every place deal and relationship data lives today. Most firms find more than they expect.
- The pipeline spreadsheet used in the weekly meeting.
- The old CRM, if there is one, including custom objects nobody remembers creating.
- Personal spreadsheets of bankers, advisers and intermediaries kept by BD or partners.
- Contacts stored only in individual Outlook address books.
- Shared drive folders per deal, with NDAs, CIMs and IC memos.
- Conference and event attendee lists.
For each source, record the owner, rough record count, last meaningful update, and a decision: migrate, archive read-only, or drop. Dropping is allowed. A list of contacts untouched for years is usually cheaper to rebuild from live email than to clean.
Field mapping and deduplication rules for CRM data migration
| Source column | Target field | Transform |
|---|---|---|
| Company | Company.name | Trim, standard legal suffixes removed |
| Website | Company.domain | Strip protocol and paths; used for matching |
| Status | Deal.stage | Map free text to your pipeline stage list |
| Lead | Deal.owner | Initials to user account |
| Source / Banker | Deal.source_contact | Match to contact by email, else create |
| Enterprise value | Deal.enterprise_value | Numeric, one currency field |
| Comments | Activity note | One dated note, author set to owner |
Deduplication rules to agree before the first load:
- Contacts match on email address first, then on name plus company.
- Companies match on domain, except shared domains, then on cleaned name.
- When two records conflict, the most recently updated value wins, and the loser is kept in history.
- A person with two employers becomes one contact with a dated job history.
- Uncertain matches go to a review list rather than merging silently.
If the new CRM fills itself from email, ask the vendor how it handles duplicates after go-live. Agree the rules anyway, so migrated data and captured data follow the same logic.
History, attachments, cutover and validation
History. Decide a lookback window. Active and recently passed deals come across in full. Older deals can come as a closed record with a single summary note. Email history may be rebuilt by capture if the new CRM reads past mail; ask the vendor how far back the first sync goes.
Attachments. Link to documents in your existing store rather than copying gigabytes into the CRM, unless the CRM is meant to be the document system. Keep access rules on restricted deal folders.
Cutover. Pick a freeze date. Do a full load a week before, a delta load on the freeze date, then set the old system to read-only. Run the next pipeline meeting from the new CRM.
Validation checks:
- Record counts reconcile by object and by stage.
- Every active deal has an owner, a stage and a last-activity date.
- Partners spot-check their twenty most important relationships.
- A random sample of records is compared field by field with the source.
- Restricted deals are visible only to their teams.
A six-week migration plan
Illustration: A made-up 12-person lower mid-market buyout fund moving from a pipeline spreadsheet and personal contact lists.
| Week | Work | Done when |
|---|---|---|
| 1 | Inventory, owners, keep or drop decisions | Signed source list |
| 2 | Stage definitions, field map, dedupe rules | Map approved by deal lead and ops |
| 3 | Email capture switched on; sample load of one fund's pipeline | Sample reviewed by two associates |
| 4 | Fix mapping issues; full load; permissions set | Counts reconcile |
| 5 | Partner spot checks; training by role | All partners signed off |
| 6 | Delta load, old sheet read-only, first meeting in the CRM | Meeting run from the CRM |
Questions & answers
How long does a CRM data migration take?
For a small or mid-sized firm moving from spreadsheets, about six weeks is a workable plan. Larger firms with an old CRM and custom objects need longer.
Should we migrate every old contact?
No. Migrate active relationships and recent deals. Archive the rest read-only. Contacts untouched for years are often cheaper to rebuild from live email.
Who should sign off a migration?
The deal lead or COO signs the field map, and each partner signs off their own key relationships before cutover.
What breaks most migrations?
Running the old and new systems in parallel for too long. Set a freeze date and make the old system read-only.
Sources for this entry
- Outlook mail API in Microsoft GraphHow apps read Outlook mail, relevant to rebuilding history from email.
- CanterisFilling the CRM from Microsoft 365 email.
Continue this reading path
- CRM email capture from OutlookCRM email capture from Outlook builds companies, contacts, deals and thread notes from Microsoft 365 mail. What to capture, exclude, merge and review.
- CRM adoption at investment firmsCRM adoption at investment firms depends on less typing, a pipeline meeting run from the CRM, clear value per role, a 90-day plan and measured usage.
- CRM security and access controlCRM security and access control for investment firms: Entra ID sign-in, roles, deal walls for MNPI, tenant isolation, audit logs, export and vendor questions.
- Pipeline management CRMA pipeline management CRM tracks each deal from sourced to closed with clear entry and exit criteria per stage, so reviews and reports stay accurate.
- How to choose a private equity CRMHow to choose a private equity CRM in six steps: requirements, a weighted scorecard you fill in, a demo script, a live-deal pilot, references, contract.
Cite this entry
Private Equity CRM. “CRM data migration.” Canteris Reference, published by Canteris. Updated . https://privateequitycrm.com/data-migration/