Skip to content

Guides · Switching

Switching CRMs: a checklist that assumes you are busy

Most CRM migrations fail in the same place, and it is not the export. It is the fortnight afterwards, when the new system holds a slightly wrong copy of the database and the old one is still the thing people actually open. This is the version of the checklist that assumes you have listings to run while you do it.

Last reviewed .

Decide what you are actually moving

The instinct is to move everything, and it is the single most expensive decision in the project. A database that has been running for six years contains a large number of records that were never worth having: duplicate entries from a lead source, people who bounced twice and have not been touched since 2021, and a long tail of names nobody in the office can identify.

Moving them costs you twice. Once in the effort of getting them across, and again every day afterwards, because a list you do not trust is a list you stop working. The useful question is not "what can we bring?" but "what would we re-enter by hand if it vanished tonight?" — and the honest answer is usually past clients, live pipeline, and the people who have referred you business.

Write that answer down before you export anything. It is the difference between a migration and a copy.

Get a clean export while you still have access

Do this first, and do it even if you have not finally decided to switch. An export is free, reversible and takes ten minutes; regaining access to a cancelled account to fetch one is none of those things. If your current vendor bills annually, the day you decide is the day to export, not the day the term ends.

Take the widest export the system offers — contacts, deals or transactions, notes, and activity history as separate files if that is how it comes. Then open them. An export that turns out to be missing notes is a discovery you want on day one rather than after you have cancelled.

Keep the raw files somewhere that is not the old CRM and not one person's laptop. They are the only complete copy of the relationship history your business has, for as long as the move takes.

Fix the data before it moves, not after

Duplicates, dead addresses and inconsistent stage names are much easier to fix in a spreadsheet than in a CRM, because a spreadsheet lets you sort by a column and see the whole problem at once. Sort by email, sort by phone, sort by last name — the duplicates announce themselves.

Agree the stage names before you move as well. Two people calling the same thing "Active" and "Under contract" is survivable in an old system everyone has learned to read around; it is corrosive in a new one, because it makes the pipeline report wrong on day one and the report is what convinces people the new system is worth using.

This is also the moment to decide what a contact means. If the old system filed everything under transactions, you will have people who exist twice — once as a buyer in 2022 and once as a seller in 2025 — and they need to become one person before they arrive.

Move in an order that keeps you working

Live pipeline first. Anything under contract or in active conversation goes across immediately, because that is the work that cannot wait for the project to finish. It is also a small enough set that entering it carefully is realistic.

Past clients second. They are the highest-value records you own and the ones most likely to be mangled by a bulk move, so they deserve the attention that the third group does not.

Everything else last, or never. If the long tail has not moved after a month, that is information: it was not worth moving.

Set one date and actually stop

The failure mode is the double-run — a fortnight where both systems are half-true, follow-ups get logged in whichever one happened to be open, and neither can be trusted. Two weeks of that will do more damage to a database than the migration itself.

Pick a date, tell everyone, and after it the old system is read-only. Keep the export and the old login for a quarter so you can go back for something you missed, but stop writing to it on the day you said you would.

What this looks like moving to ProspectKeeper

Worth being straight about the current state: contacts come across from a spreadsheet, while deals and follow-ups are entered in the app. For a live pipeline and a past-client list that is a focused session or two rather than a migration project.

So if your list is large, write to hello@prospectkeeper.com before you commit and we will tell you where import stands rather than letting you find out in week two. That is a worse answer than "upload your file", and it is the true one.

What you get in exchange is a database organised around the client rather than the transaction — a person can exist with no deal at all, a closed sale leaves the file open, and a second transaction on the same person takes the next filing letter beside the first. Which means the thing you were most afraid of losing in the move is the thing the system is actually built to keep.

The checklist

Copy it out. The order matters more than the items.

  1. Export everything from the current CRM today, before any decision is final.
  2. Open the export files and confirm notes and activity history are actually in them.
  3. Store the raw export somewhere that is not the old CRM and not one laptop.
  4. Decide what you would re-enter by hand if it vanished — that is the real migration scope.
  5. De-duplicate in the spreadsheet by email, then phone, then surname.
  6. Agree stage names across the team before anything moves.
  7. Merge people who exist twice because the old system filed by transaction.
  8. Move the live pipeline first, and carefully.
  9. Move past clients second — they are the records worth the most.
  10. Move the long tail last, or accept that it was never worth moving.
  11. Set a cutover date and make the old system read-only after it.
  12. Keep the old login and the export for one quarter, then let it go.

Questions

When is the right time to export?

The day you start seriously considering a move. An export costs ten minutes and nothing else, and it removes the worst pressure from the decision — you are no longer negotiating with a vendor who is holding the only copy of your relationship history.

Should I move my whole database?

Usually not. Past clients, live pipeline and the people who have referred you business are almost always worth moving; a six-year tail of unidentifiable names generally is not. A list you do not trust is a list you stop working, so bringing everything across can cost you more than leaving it.

Can I import a CSV into ProspectKeeper?

Contacts come across from a spreadsheet: download the template, map your columns and bring the list in. Deals and follow-ups are entered in the app afterwards, which for a live pipeline is a session or two. For a very large or messy database, write to hello@prospectkeeper.com first and we will help you plan the move.

How long should a switch take?

Plan for a fortnight of overlap at most, with a hard cutover date inside it. The risk grows with the length of the overlap rather than the size of the database — two systems that are each half-true for a month will damage a list more than the move itself does.

What usually gets lost?

Notes and activity history, because they are the parts exports handle worst and the parts nobody checks until they need them. Open the export files on day one and look specifically for those two, rather than confirming the contact count and assuming the rest came with it.

What if I am leaving the brokerage as well as the CRM?

Treat them as two separate problems and do the export before you give notice, because access tends to end on the day the conversation happens rather than the day you leave. What you are entitled to take is set by your agreement and by state law, so read it before you plan around it.

See all questions

Read next

Try it against your own pipeline

Free to start. Set up your company, add your clients and work the list.

Start freeMore guides