How-to guides

How to migrate to a new CRM without losing data or momentum

A migration is won or lost before the import button is pressed — in the cleanup and the mapping. Here is a plan you can follow with any CRM, and the exact import screens NeuroCRM gives you for it.

In short: A CRM migration has eight steps: decide what moves, take inventory, export and clean the data, map old fields to new ones, build the target and run a small test import, cut over on a set day, train the team, and keep the old system read-only for a while. In NeuroCRM, contacts and deals come in from CSV or Excel files (Contacts → Import, or Settings → Import / Export); only Bitrix24 has a direct transfer. There is no live connector from HubSpot or Salesforce, so those move by file.

What does a CRM migration actually involve?

You are moving three things, and only one of them is data: records (contacts, companies, deals, notes), structure (pipelines, stages, custom fields, tags, roles) and habits (where people log a call, who owns a lead, what happens when a deal stalls).

Treat it as a small project with an owner, a date and a rollback plan, not a file transfer.

The CRM migration checklist

Work through the steps in order. Each has a "done when" test.

1. Decide why you are moving and what comes along

Write down the two or three reasons: cost, a missing channel, a tool nobody opens. Then set the scope. Done when you have a one-page list of what moves, what stays archived, and who signs off.

2. Take inventory of the old system

List every object and setting that someone relies on: contacts, companies and deals with rough counts; pipelines and stage names; custom fields (who fills them, whether anyone reads them); tags and lists; users and record owners; automations, templates, forms and integrations that write into the CRM; attachments, call recordings and email history.

File exports rarely carry the last group, so find out now what the old system can and cannot export. Done when every item is marked "move", "rebuild by hand" or "leave behind".

3. Export and clean before you import

Export what you marked "move" to CSV or Excel and keep an untouched copy. Then clean the working copy:

  1. Merge duplicates. One person under two emails means two people follow up, or nobody does.
  2. Delete the dead: test records, old leads nobody ever called, contacts with no way to reach them.
  3. Normalize phones and emails. One phone format, lowercase emails, no "n/a" in the email column.
  4. Fix names. One pattern, with the company in its own column.
  5. Standardize stage and source values. "Won", "won" and "Closed Won" become one value, because importers match text literally.
  6. Remove columns nobody uses. Each one is a field you must map and maintain.

Done when a colleague scans fifty random rows and finds nothing embarrassing.

4. Map old fields to new ones

Make a mapping table before you touch the importer.

Old fieldNew fieldNotes
Full nameNameRequired in most importers
Work emailEmailAlso the key used to spot duplicates
MobilePhoneSame format as the rest
Lead sourceSourcePick one list of values
Deal stageStageMust match a stage that already exists in the new pipeline
Owner(assign after import)Check whether your importer has an owner column
Anything elseCustom fieldOnly if someone will actually use it

Mark each row "direct", "needs transformation" or "drop". Transformations are where migrations stall: a status field with 14 values that must collapse into six stages needs a person's decision, not a formula. Done when no column lacks a destination or a "drop" mark.

5. Build the target before you load into it

Create pipelines, stages, tags and custom fields in the new CRM first, because an importer can only match a stage name to a stage that exists. Decide who the users are and what each role sees.

Keep automations switched off while you load data. Creating records usually triggers rules, and you do not want a welcome message going out to thousands of old addresses.

6. Run a test import with a small slice

Import 50 to 100 rows that include your messy cases: a duplicate, a record with no email, a long note, a deal in an odd stage. Open ten imported cards and compare them with the old system line by line. Fix the file or the mapping, import the slice again, and only then load everything. Done when two people agree the test cards match.

7. Cut over on a fixed day

  1. Announce the freeze: from this evening nobody edits the old CRM.
  2. Export the delta (records created or changed since your test export) and import it.
  3. Compare counts of contacts, deals and open deal value with the old system.
  4. Switch on automations and connect your channels.
  5. Tell the team, and whoever runs the forms that fed the old system, where the new one lives.

8. Train people and keep the old system read-only

Give each role a 30-minute walkthrough of the three actions they do daily, not a tour of every menu. Keep the old CRM read-only for a set period, such as a month or a full sales cycle, so "where did that quote go?" has an answer.

Do you need a migration service or a tool?

Services and tools exist for large moves: many custom objects, years of email history, several source systems. They cost money and still need your mapping decisions. A smaller business moving contacts and deals can often do it alone with the steps above. Whoever you hire, ask for a test import on your own data and for a list of what they cannot move.

How does migration work in NeuroCRM?

Files bring in contacts and deals through two routes that behave differently; Bitrix24 has a direct transfer that also brings companies.

Contacts → Import

On the Contacts page, the Import button opens a file dialog. It reads CSV, TSV, Excel (.xlsx, .xls, .xlsm, .xlsb) and OpenDocument (.ods) files and recognizes the delimiter and UTF-8 encoding on its own. You then:

  • tick First row is headers if your file has them;
  • see a preview of the first five rows;
  • map each column to a standard field (Name, Email, Phone, Company, Position, Source, Tags, Notes), skip it, or send it to a Custom field;
  • if you are an admin, AI mapping suggestion proposes fields for unrecognized columns from their sample values (it uses your workspace's AI key or credits) — review it before you accept it;
  • optionally type a list name, which creates a static list from the file and gives every imported contact a tag with the same name.

One column must be marked as Name. A Download CSV template link gives you a file in the expected shape.

A row whose email or phone already exists updates that contact instead of creating a second one, and the result screen counts created and updated records separately. Fields already filled in on the card are not overwritten; the dialog lists the file rows that differ so you can decide by hand.

A Company column links the contact to a company; companies come in through the contacts that mention them.

The History button on the Contacts page lists past file imports with their status and a Roll back button (admins) that deletes the contacts that batch created; contacts it only updated stay as they are. If some already have linked data, you are asked to confirm first. That makes the test-import step low-risk.

Settings → Import / Export

This is the admin route for contacts and for deals, with CSV or XLSX files up to 20 MB. Drop a file on the card and the result shows how many rows were new, updated and skipped, plus row-level errors and the columns it detected.

  • Contact columns: name, email, phone, company, position, tags, source, notes.
  • Deal columns: title, amount, currency, contact email, contact phone, contact name, stage, tags, won.

Headers are recognized by name, including variants such as "deal name", "value" or "status"; rename a column if its field is not detected. Rules for deals:

  • Stage is matched by name against existing stages, ignoring case. A name that matches nothing, or an empty one, puts the deal into the first stage of your first pipeline.
  • At least one pipeline must exist. Without one, every row comes back as an error.
  • The contact is found by email, then phone. If neither matches and the row has a contact name, a contact is created on the spot.
  • A repeat of the same title for the same contact overwrites that deal with the file's stage (blank or unknown means the first stage), amount and won value (no won column means not won); a row with no matching contact becomes a new deal every time.
  • There is no owner or created-date column, so plan to assign owners after the import.

The same screen has Download CSV buttons for contacts and deals. Use them to take a backup of your new CRM right after the final import.

Bitrix24: direct transfer

If you are leaving Bitrix24, the section "Migration from Bitrix24" on the same Import / Export screen connects to your portal with an inbound webhook URL. It offers a trial migration of 20 records, a full migration that runs in the background, resume, and roll back, and it lists what it does not transfer before you start. Read that list.

Moving from HubSpot, Salesforce, Pipedrive or others

There is no live connector for these: export a file and use the routes above. The NeuroAdmin chat can also take an attached file and recognizes HubSpot and GoHighLevel exports by their English column headings, but it is not enabled in every workspace.

A NeuroCRM-specific cut-over checklist

  1. Create pipelines and custom fields before importing.
  2. An industry setup only adds items, and its automation rules are created switched off. Leave them off until the import finishes.
  3. Import contacts first, then deals, so deals can find their contacts.
  4. After cut-over, check which pipeline is first. Deals created automatically from new chats, and from web forms with a field marked "Save to contact", land there, without an owner.
  5. Run your test import inside the 14-day trial, which includes everything and has unlimited users; after it ends the workspace is read-only. See pricing.

What NeuroCRM does not do for a migration

  • No live connector from HubSpot, Salesforce, Kommo, Pipedrive or similar systems. They move by file. Only Bitrix24 has a direct transfer.
  • Files carry contacts and deals only. There is no separate file import for companies, so companies come from the Company column of a contacts file.
  • The NeuroAdmin chat import is not enabled in every workspace, so do not build your plan on it before checking.

If your old CRM holds years of custom objects or email threads that must stay linked to records, plan that archive separately and keep the old system read-only.

FAQ

How long does a CRM migration take?

It depends on cleanup, not file size. Loading a clean file takes minutes; deciding what to keep, merging duplicates and agreeing on stages takes the real time.

Can I migrate from HubSpot or Salesforce to NeuroCRM?

Yes, by file. Export contacts and deals as CSV or Excel, map the columns, test with a small slice, then load the rest.

What can I import into NeuroCRM?

Contacts and deals. Contacts go in through Contacts → Import or Settings → Import / Export; deals through Settings → Import / Export, or as a file attached in the NeuroAdmin chat where that is enabled.

Can I undo an import?

Partly. Files loaded through Contacts → Import can be rolled back by an admin from History, which deletes the contacts that run created. For the Settings route, test on a small file and keep your export.

Will imported records trigger my automations?

New contacts and deals are events that automation rules can react to, so a rule you have already switched on can fire for every imported record. Keep rules off until the import is done.

How much does NeuroCRM cost after the migration?

$13 per user per month, or $10.40 billed yearly. Test the import first in the 14-day trial; see the pricing page.

How this article was made: written by the NeuroCRM team with AI assistance; product claims were checked against NeuroCRM code and help docs; prices are from the published pricing as of the update date.