Nearly every CRM migration post-mortem blames the data. Nearly every CRM migration actually fails on adoption.
Records transfer reliably; modern tooling has made that part close to solved. What does not transfer is the reason a field was filled in a particular way, the stage vocabulary the team had internalised, and the reporting someone in finance depended on without telling anyone.
What to migrate
- Accounts, contacts and open deals. Non-negotiable.
- Closed-won history for the last 24 months. You need it to compute conversion rates and dwell times, which is how the new pipeline gets designed.
- The engagement timeline for anything currently open. A rep opening a live deal on day one and seeing no history will not trust the system again.
What to leave behind
- Closed-lost records older than two years. They inform nothing and they make every list slower.
- Custom fields with under 20% fill rate. If nobody populated it, it was not load-bearing.
- Automations you cannot explain. Rebuild them deliberately or discover you did not need them.
The 20% rule: run a fill-rate report on every custom field before migrating. Anything under 20% populated is a field somebody added once and nobody maintained. Migrating it carries the clutter forward and slows adoption of the new system.
The three failures that actually happen
- Vocabulary mismatch. The old system said "project" and the new one says "deal". Adoption drops immediately and quietly, and nobody reports it as a migration problem.
- Silent reporting dependency. Someone in finance had a saved report feeding a board pack. Nobody knew. It breaks at quarter end.
- Dual running for too long. Two systems live simultaneously means neither is trusted. Set a hard cutover date and enforce it.
What each industry has to preserve
Vocabulary and stages to preserve, by industry
| Industry | Pipeline stages | Records renamed |
|---|---|---|
| Advertising | Lead → Brief → Pitch → Client Approval → Live | Deal→Campaign, Account→Brand |
| Construction | Lead → Site Survey → Estimate → Tender → Awarded | Deal→Project, Account→Client |
| Consulting | Enquiry → Scoping → Proposal → Commercials → Engaged | Deal→Engagement, Contact→Client, Account→Client Org |
| Cybersecurity | Prospect → Discovery → Demo → Trial/POC → Negotiation → Closed Won | — |
| E-commerce | Enquiry → Quote → Order Placed → Fulfilled → Repeat | Deal→Order, Contact→Customer |
| Education & EdTech | Enquiry → Counselling → Offer → Fee Payment → Enrolled | Deal→Enrolment, Contact→Student, Account→Institution |
| Financial Services | Lead → KYC → Credit → Sanctioned → Disbursed | Deal→Application, Contact→Applicant, Account→Employer |
| Healthcare | Enquiry → Consultation → Insurance Approval → Treatment → Closed | Deal→Case, Contact→Patient, Account→Referrer |
| Insurance | Enquiry → Quotation → Underwriting → Policy Bound → Renewal | Deal→Policy, Contact→Policyholder |
| Legal Services | Enquiry → Conflict Check → Scoping → Engagement Letter → Active Matter | Deal→Matter, Contact→Client, Account→Organisation |
| Logistics & Freight | Enquiry → Rate Quoted → Booking Confirmed → In Transit → Delivered | Deal→Shipment, Account→Shipper |
| Manufacturing | Enquiry → RFQ → Costing → Sampling → PO Received | Deal→Order, Account→Buyer |
| Marketing Agency | Lead → Brief → Pitch → Client Approval → Live | Deal→Campaign, Account→Brand |
| Non-profit | Prospect → Cultivation → Proposal → Under Review → Funded | Deal→Grant, Contact→Donor, Account→Funder |
| Real Estate | Enquiry → Site Visit → Negotiation → Booking → Registered | Deal→Listing, Contact→Buyer, Account→Developer |
| Recruitment | Enquiry → Scoping → Proposal → Commercials → Engaged | Deal→Engagement, Contact→Client, Account→Client Org |
| SaaS & Software | Prospect → Discovery → Demo → Trial/POC → Negotiation → Closed Won | — |
| Travel & Hospitality | Enquiry → Itinerary → Advance Received → Confirmed → Travelled | Deal→Booking, Contact→Guest, Account→Agency |
The sequence that works
Design the pipeline from historical conversion data before you move anything. Map vocabulary to what the team already says. Migrate open work and 24 months of history. Set a cutover date. Turn the old system read-only on that date rather than deleting it.
Migrations that start with data transfer and design the pipeline afterwards recreate the old system's problems in a new interface.
Quotarider ships all 18 industry packs. Records renamed, stages preconfigured, and the chain followed from first touch through to the paid invoice — so you can see which activity produced collected revenue rather than bookings.
Start free in 5 minutes →Frequently asked questions
How much CRM history should I migrate?
Open deals and their full engagement timeline, plus roughly 24 months of closed-won history so you can compute conversion rates and dwell times. Closed-lost records older than two years rarely justify the clutter.
Why do CRM migrations fail?
Usually adoption rather than data. The most common causes are vocabulary that no longer matches how the team speaks, an undiscovered reporting dependency that breaks at quarter end, and running two systems in parallel long enough that neither is trusted.
Should custom fields be migrated?
Run a fill-rate report first. Fields under about 20% populated were added once and never maintained; carrying them forward slows adoption of the new system without preserving anything useful.
How long should the old CRM stay available?
Set a hard cutover date, then make the old system read-only rather than deleting it. Indefinite dual running is the most reliable way to ensure neither system is trusted or maintained.
Sources: Google Search Central, Google bulk sender guidelines, and Quotarider industry pack configurations.