The CRM Migration Checklist

Before You Commit to Anything

Inventory first. You cannot get a real quote or a real timeline without these numbers, and gathering them takes an afternoon.

Count your contacts and companies. Count your open deals and your closed ones separately, because you may not need the closed ones. Count your pipelines. Count your live automations, and be honest about which are genuinely running.

Then the one people forget: list every system that holds customer data. Your CRM, your email platform, your scheduler, your forms tool, your support inbox. Each one is a potential migration inside the migration, and this is routinely the piece nobody prices for.

Finally, decide your cutoff. Five years of dead contacts is not an asset, it is weight. Pick a date and be willing to leave things behind.

Clean Before You Move

Dirty data that gets migrated is just dirty data somewhere new, and it is far cheaper to fix in a spreadsheet than inside a live system.

Export and look at it properly. Phone numbers will be in several formats. Dates will not parse consistently. There will be email addresses with obvious typos, names in capitals, and duplicate records that have been quietly breeding for years.

Fix all of that in the export file. It is tedious and it is the highest-return hour of the whole project, because every one of those problems becomes permanent the moment it lands in the new system and starts getting referenced.

Deduplicate last, after the formatting is consistent, because inconsistent formatting hides duplicates.

The Order to Migrate In

Order matters more than people expect, because each step depends on the one before it.

Pipelines and custom fields first. The containers have to exist before anything can go into them.

Contacts and companies second. This is the reliable part and it gives you something to attach everything else to.

Deals third, with owners and values explicitly mapped. Depending on the tools involved, values sometimes need a second pass after the deals land, so plan for that rather than discovering it.

Automations last, rebuilt rather than moved, and switched off until you have verified everything else. Turning sequences on before the data is verified is how prospects get emailed twice.

Then, and only then, email archive as a separate exercise.

Test With Ten, Not Ten Thousand

This is the step people skip under time pressure, and it is the most expensive shortcut available.

Migrate ten records first. Pick deliberately: one simple contact, one with a long history, one deal at an early stage, one deal about to close, one record you know is messy.

Then check them properly. Did ownership come across. Are deal values attached. Is the deal in the right stage. Did custom fields land in the right places. Does the contact history look right.

Fix whatever is wrong in the mapping, then run the ten again. Only when the sample is clean do you run the full load.

Finding a mapping error after eighteen thousand records have landed means undoing eighteen thousand records, and that is a bad week.

Verify Before You Trust It

Before you tell the team the new system is live, run these checks.

Spot check twenty records against the old system. Not counts, actual records, opened side by side.

Confirm every open deal has an owner. Run a filter for deals with no owner. It should return nothing. It frequently does not.

Send one real test through every rebuilt sequence and confirm it arrives, from the right address, with working links.

Check what your team sees rather than what you see. Admin views hide permission problems.

And confirm your reporting returns numbers you recognise. If the new dashboard says something different from what you know to be true, find out why before anyone starts making decisions on it.

The First Thirty Days After

Go-live is the middle of the project, not the end. This is the part almost everybody underestimates.

Turn the old system read-only rather than deleting it, and keep it for at least ninety days. You will need it, usually for something small and urgent.

Watch for silence. The most common post-migration failure is an automation that stopped without anyone noticing. Check that your sequences are actually sending in week one, not just that they exist.

Run a short daily standup for the first two weeks. Not a meeting about the CRM, a meeting where the pipeline is on screen. That single habit does more for adoption than any amount of training.

Expect a dip. Productivity drops for a few weeks while everyone relearns where things live. Plan for it rather than panicking about it.

And after thirty days, ask your team the honest question: is this better? If the answer is no, something in the design is wrong and it is much cheaper to fix now than in a year.

Before you commit. Count contacts, companies, open and closed deals, pipelines, live automations, and every system holding customer data. Decide your data cutoff date.

Before anything moves. Write down your stages and owners. Build the pipelines in the new system first. Clean the export in a spreadsheet.

During. Migrate ten records, verify ownership and values, then run the full load. Contacts and companies, then deals, then automations.

Before go-live. Spot check twenty records. Confirm every deal has an owner. Send one test through every rebuilt sequence.

After. Old system read-only for ninety days. Daily standup for two weeks. Ask the team at thirty days whether it is actually better.