How to migrate CRM software
The switching cost nobody prices in: the CSV moves your contacts and deals, but the activity timeline, attachments, automations and every integration do not — and the real project is getting reps to log in and trust the new pipeline. What breaks, what to protect, and the sequence that survives a live sales team.
The rows move; the system doesn’t
Every CRM migration write-up fixates on exporting contacts. That is the easy half. Contacts, companies and deals leave as clean CSVs. What does not leave is everything that made the CRM useful: the activity timeline on each record, the logged calls and emails, the attachments, the pipeline automations, the lead-routing rules, the sequences, the reports and dashboards, the role and sharing permissions — and every integration, all of which point at the old system and re-authenticate from scratch.
The deeper risk is adoption. A CRM is worth exactly the share of your reps who actually log activity in it, and a migration is the moment that share can crater. Run two systems too long and the pipeline splits in half; rush the cutover with broken automations and reps quietly revert to a spreadsheet. Salesforce makes both halves worse, because its custom objects, flows, validation rules and sharing model — not the records — are the value, and none of it exports. Treat this as a change-management project, sequence it deliberately, and anchor it to what you chose in how to choose CRM software.
The five phases of a CRM migration
List your objects and record counts, custom fields and their types, pipelines and stages, automation and lead-routing rules, sequences, reports, dashboards and permission roles. Then list every connected system: email and calendar sync, marketing automation, telephony, e-signature, quoting, accounting, and any Zapier or native connectors. Naming the non-exportable logic and the integrations now is what prevents a silent break after cutover.
Migration is the one moment you get to leave the junk behind, so de-duplicate contacts and merge stale records before export, not after. Choose a data cutoff, export contacts, companies and open deals to CSV, and decide deliberately how much activity history to carry — most teams keep the old CRM read-only for the timeline rather than forcing notes and emails through an importer that mangles them.
Recreate pipelines and stages, custom fields and objects, automation and lead-routing rules, sequences, reports, dashboards and permission roles by hand in the new CRM — none of it ports, and on a Salesforce exit this is the bulk of the work. Map every old field to a new one explicitly. Get the scaffolding right before you load a single live deal, or you will rebuild it with the pipeline already running.
Import records against your field map and verify the de-dupe held. Then rebuild and re-authenticate every integration — email/calendar sync, marketing, telephony, e-sign, quoting, accounting — and confirm each trigger fires. Run a real pilot: one sales team works fully in the new CRM for a couple of weeks while the rest finish in the old one. The pilot exposes the dead automation and the lead form that stopped creating contacts before the whole org depends on them.
Set the cutover date up front and hold it. On the date, move everyone, switch the old CRM to read-only so history stays reachable, and run short, role-specific retraining so reps’ logging habits transfer with the data. Monitor adoption and data hygiene for the first month — open rates of the new pipeline, activities logged per rep. Indefinite parallel running is how CRM adoption dies; a clean, dated cutover is what makes it stick.
The export gotchas, named
Mostly no, and it is the loss teams discover after cutover. Contacts, companies and deals export to CSV as field data. The timeline wrapped around each record — logged calls, email threads, meeting notes, task history and file attachments — generally does not travel, or travels only partially through a paid migration tool. If that history is your relationship memory, plan to keep the old CRM read-only for reference, or use a purpose-built migration service that explicitly moves activities and notes. Never assume a CSV import carries the story behind the deal.
You can move the records, but not the system. Standard and custom objects export, yet the things that make a Salesforce org valuable — custom objects with complex relationships, validation rules, flows and Apex, page layouts, roles and sharing rules, and report types — have no portable format and are rebuilt by hand in the new tool. This is why migrating off Salesforce is the hardest move in the category: the configuration is the value, not the rows. Budget most of the project for rebuilding logic, and use exports only for the raw data.
For a small team with a light setup — a few thousand contacts, a simple pipeline, few integrations — one to two weeks. For an organization with custom objects, automations, sequences, connected marketing and telephony, and role-based access, four to ten weeks is realistic, and a Salesforce exit can run a quarter or more. The data load is rarely the bottleneck; rebuilding the automations, re-authenticating integrations and getting reps to actually log in are. It is a change-management project wearing a data-migration costume.
No integration comes across automatically — every one is re-established and re-authenticated in the new CRM. Email and calendar sync, marketing automation, telephony and dialers, e-signature, quoting/CPQ, accounting, and any Zapier or native connectors all point at the old system and must be rebuilt and retested. This is the most common source of silent breakage: a lead form that stops creating contacts, or a sequence that quietly stops sending. Inventory every integration up front, and test each trigger with a pilot team before the full cutover.
Briefly, with a hard cutover date — never open-ended. A short pilot where one sales team works fully in the new CRM while the rest finish in the old one surfaces broken automations and missing fields safely. But indefinite parallel running is where CRM migrations die: contacts and deals get updated in two places, nobody trusts the pipeline numbers, and reps drift back to spreadsheets. Set the cutover date up front, switch the old CRM to read-only on it, and retrain.
One email when the rankings move. The shortlist, the tradeoffs, the price changes. No filler.