Switching CRM takes days for a simple move and months for a customised Salesforce instance. Contacts and open deals transfer cleanly. Activity history transfers badly. Automations, dashboards and reporting continuity do not transfer at all. Budget for partner fees, a productivity dip, and rebuilding every workflow by hand.

What transfers, and what does not

Transfers cleanly

  • Contacts and companies
  • Open deals and pipeline stages
  • Custom fields on standard objects
  • Notes

Transfers badly

  • Activity history
  • Email threads
  • Attachments and files
  • Custom object relationships

Does not transfer

  • Automation and workflow logic
  • Reports and dashboards
  • Historical reporting continuity
  • Integrations and their configuration

The last group is the one that hurts.

Every automation you have built gets rebuilt by hand. Lead routing, stage-change triggers, task creation, notification rules: none of it exports, because each platform expresses that logic differently. A workflow that took an afternoon to design takes another afternoon to recreate, and you will have more of them than you remember.

Reporting continuity is the second surprise. Your year-on-year trend data restarts on switch day. You can import historical records, but the reporting layer will not reconstruct what your pipeline looked like last March. Leadership loses the comparison right when they most want it.

Real timelines

  • Pipedrive to HubSpot, 10 reps, simple pipeline: days. Both sides have native import tools and the data models are similar.
  • HubSpot to Attio, 20 reps, some custom objects: two to four weeks, most of it spent rebuilding automations and remapping objects.
  • Salesforce to anything, 50 reps, real customisation: three to six months, and you will pay a partner. Record transfer is the cheap line, quoted in the low five figures by migration services. Rebuilding logic, integrations and reporting is the engagement, and that is where the budget actually goes.
  • Anything to Salesforce: budget for longer than the quote says. Customisation, validation rules and integrations are what stretch it, so the gap is widest on instances that have been shaped over years.

The variable is not record count. It is how much logic sits on top of the records. A hundred thousand contacts move in an afternoon. Forty validation rules and a dozen custom objects do not.

The costs that never appear in the quote

The productivity dip. Your reps relearn a tool while carrying a number. Expect reduced output across the team for several weeks. It rarely appears in a migration plan and it is usually the largest cost in one.

Adoption reset. Whatever CRM hygiene you had built up, you rebuild. Some of it never comes back, because the habits were tied to the old interface.

Data cleanup you did not plan for. Migration surfaces every duplicate, every half-filled field and every deal nobody closed out. You either fix it before importing, which takes time, or you carry it into the new system, which wastes the clean start. Most teams underestimate how much of this there is.

Rebuilt integrations. Every connected tool needs reconnecting and reconfiguring, and the ones held together with Zapier need rethinking entirely.

Running two systems at once. Most teams keep the old CRM live for reference well past the switch, so budget for overlapping subscriptions. Annual contracts make this worse, because cancelling mid-term usually means keeping access rather than getting money back.

Which platforms are hardest to leave

Switching cost is a buying criterion, and almost nobody treats it as one.

  • Hardest: Salesforce. Not for data reasons but for logic. Validation rules, flows and Apex represent years of accumulated decisions, and none of it exports.
  • Hard: HubSpot, if you use marketing alongside sales. The shared contact record that made it valuable is what makes it sticky.
  • Moderate: Attio, monday, Close. Modern APIs and clean exports, but custom objects and automations still get rebuilt by hand.
  • Easiest: Pipedrive, Zoho, Copper, Folk. Simpler data models mean less to lose.

How to make it hurt less

Do not migrate everything. The instinct is to bring all of it. Resist. Move open deals, active accounts and contacts from the last 18 to 24 months. Archive the rest as a CSV export you can search if you ever need it. Most teams never do.

Export before you cancel. Obvious, and people still get caught. Most platforms retain data for a limited window after cancellation, then delete it permanently. Take a full export while the account is live, including attachments.

Run parallel briefly, not indefinitely. Two weeks of overlap catches what you missed. Two months of overlap means nobody commits to the new system.

Rebuild automations last. Get records and pipeline in first and let people work. Automations built before you have seen the new tool in use tend to be recreations of old habits rather than improvements.

Pick a quiet moment. Not quarter end. Not the week a big deal closes. There is no perfect time, but there are clearly bad ones.

Assign one owner. Migrations that belong to everyone belong to nobody. One person decides what moves, what gets archived and when the old system goes dark.

When switching is worth it anyway

Migration is survivable. Thousands of companies do it every year and come out better, and fear of the process keeps more teams on the wrong CRM than the process itself has ever cost. If reps route around your CRM, or leadership does not trust the numbers coming out of it, a few weeks of disruption is cheaper than three more years of working around it. The point is not that switching is a disaster. It is that switching is expensive enough to be worth avoiding once, at the moment you choose, by picking for where you are going rather than where you are.

The best CRMs for B2B sales teams in 2026 →

Frequently asked questions

A simple move between similar tools, such as Pipedrive to HubSpot with ten reps, takes days. Moving twenty reps with custom objects takes two to four weeks. Leaving a heavily customised Salesforce instance takes three to six months and usually a partner. Record count matters less than how much logic sits on top.
Contacts, companies, open deals and notes transfer cleanly. Activity history, email threads and attachments transfer badly or partially. Automations, dashboards, reports and integration configurations do not transfer at all and must be rebuilt by hand. Historical reporting continuity is lost regardless of how much data you import.
Simple self-serve migrations cost nothing but time. Partner-led migrations split into two lines: record transfer, quoted in the low five figures by migration services, and the engagement that rebuilds logic, integrations and reporting, which is the larger of the two. Larger still are the costs nobody quotes: reduced rep productivity for several weeks, a reporting blackout during transition, and a period of paying for both systems.
Usually not. Move open deals, active accounts and contacts from the last 18 to 24 months. Export everything else as a CSV archive you can search if needed. Most teams never open it. Migrating everything slows the project and imports data quality problems into a clean system.
Salesforce, by a distance. The difficulty is not the data but the logic: validation rules, flows and custom code represent years of decisions and none of it exports. HubSpot is next hardest if you use marketing alongside sales. Pipedrive, Zoho, Copper and Folk are the easiest to leave.

Timelines and cost indications here are drawn from published implementation guidance rather than vendor-published rates, and they vary widely with instance complexity. Last reviewed September 2026.