Lizzard Solutions

Data Migration Checklist: Switch Systems Without Data Loss

· Lizzard Solutions
Data MigrationCRMData ManagementSmall Business ITChecklists

By Nemanja Vukmirović, Lizzard Solutions. Last updated 17 November 2026.

You picked a new CRM, accounting tool or online shop. Now the hard part: moving years of customers, invoices and products without losing records, breaking reports or leaving staff with two half-working systems.

This data migration checklist walks a small business through the whole move in four phases: plan, prepare, move and verify. You get a copy-ready checklist, a field mapping table, a test plan, a rollback rule and a worked example, so the switch happens on one planned day instead of dragging on for months.

TL;DR

  • Data migration is moving business data from an old system to a new one, with the data cleaned, mapped to new fields and checked after the move.
  • Back up the old system first, map every field, run a test migration on a copy, and only then do the real move on a fixed cut-over date.
  • Verify with record counts and spot checks, keep the old system read-only for a set period, and only then switch it off.

What is data migration?

Data migration is the process of moving data from one system to another, for example from spreadsheets to a CRM or from an old accounting program to a new one. A good migration includes a backup, cleaning, mapping old fields to new ones, a test run, the final move and a verification that every important record arrived correctly.

Most failed migrations are not technical disasters. They are small gaps: phone numbers cut off, a missing customer type, dates read in the wrong format, or attachments left behind. A checklist catches those before staff and customers do.

Key takeaway

Never migrate without a tested backup of the old system and a written list of what "done" means. Those two things make every mistake fixable.

Action for today: write one sentence describing what moves, from where, to where, and by which date.


Phase 1: plan the migration

  1. List the data sets. Customers, contacts, deals, products, invoices, documents, users. Your IT asset inventory template helps you find every system involved.
  2. Decide what not to move. Old, duplicate or unneeded records stay in the archive. Personal data you no longer need is a liability, not an asset.
  3. Name an owner per data set. The person who uses the data daily decides what correct looks like.
  4. Pick the cut-over date. Choose a quiet period, not month-end or a sales peak, and tell the team early.
  5. Write the definition of done. For example: all active customers moved, open invoices match totals, five users trained.

Action for today: mark the data sets you will not migrate and note where the archive will live.


Phase 2: back up, export and clean

  1. Take a full backup of the old system and test a restore. The NCSC small business guidance on backing up data recommends identifying essential data, keeping backups separate from your main systems and making backing up a routine. Before a migration this is not optional.
  2. Export to an open format such as CSV or XLSX, plus attachments in folders. Check that the export includes every column, not only the visible ones.
  3. Clean the data in the export, not in the old system. Merge duplicates, fix obvious typos, standardise phone, date and country formats.
  4. Freeze changes to structure. From now until cut-over, no one adds new custom fields to the old system.

For a full backup routine that outlives the project, see our 3-2-1 data backup guide for small business.

Action for today: run the export and open it. Count rows per data set and write the numbers down.


Phase 3: map fields from old to new

Field mapping is where most errors hide. Create a table with one row per old field:

Old fieldNew fieldRuleExample
Customer nameCompany nameCopy as isPekara Zlatni Klas
PhonePhoneAdd +381, remove spaces+381641234567
Date (text)Created dateConvert DD.MM.YYYY to YYYY-MM-DD2024-03-15
Type (A/B/C)Customer tierA=Key, B=Regular, C=InactiveKey
NotesNote (activity)Import as one note per customerPrefers delivery Tuesday

The rows above are example data; replace them with your own fields. Every old field must have a decision: map, merge, transform or drop. An empty decision is a bug waiting to happen.

If the old and new systems both have APIs, the move can be scripted instead of done with files. Our API integration guide explains when that pays off.

Action for today: build the mapping table and mark every field without a decision.


Phase 4: test, move and verify

Run a test migration

Import a sample, ideally 50 to 100 records of every type including awkward ones, into a test account or sandbox. Ask each data owner to check their records against the old system and list what is wrong. Fix the mapping, then repeat until a test run comes back clean.

Cut-over day

  1. Announce a short freeze: no new entries in the old system from a set time.
  2. Take a last export and the final backup.
  3. Run the full import using the tested mapping.
  4. Set the old system to read-only, if it allows it, and keep its logins for the agreed period.

Verify

CheckHowPass when
Record countsCompare row counts per data set, old vs newEqual, or difference explained (dropped duplicates)
TotalsSum open invoices or balances in both systemsTotals match exactly
Spot checksEach owner opens 10 random recordsAll fields correct, attachments open
Users and accessLog in as each roleRight people see the right data

Rollback rule

Decide before cut-over what failure means, for example totals off or records missing, and who can call a rollback. With a tested backup, rolling back is a decision, not a crisis.

Action for today: write your four verification checks and the name of the person who signs them off.


Security and privacy during the move

Export files are full copies of your business data sitting outside any system. Treat them carefully: store them in an access-controlled folder, never email them as attachments, and delete working copies after verification.

If the data includes personal data, GDPR Article 32 requires appropriate security measures, and Serbia's ZZPL has an equivalent rule. Exporting data into a structured, commonly used format also mirrors the idea behind the right to data portability, so a good new vendor should make both import and export easy. Ask about export before you sign, as covered in our guide to choosing a CRM for small business.

Action for today: create one restricted folder for migration files and set a date to delete them.


Worked example: spreadsheets to a CRM

This is a fictional example to show the process, not a real client. A furniture workshop with eight staff moves customer data from three spreadsheets and an email address book into a new CRM.

  1. Plan: they move active customers and open quotes, and archive anything older than their retention policy allows.
  2. Prepare: backup taken, exports combined into one file, duplicates merged by phone number.
  3. Map: twelve old columns get a decision; two are dropped, one is split into city and address.
  4. Test: a 60-record test import shows dates read in the wrong format; the rule is fixed and the second test is clean.
  5. Move and verify: cut-over on a Friday afternoon, counts and spot checks pass on Monday, and the spreadsheets become read-only for 60 days.

Action for today: pick your own quiet day for cut-over and block it in the calendar.


Frequently asked questions

How long does a small business data migration take?

For one system with clean data, such as spreadsheets into a CRM, a few days of preparation and one cut-over day is common. Several systems, messy data or attachments can stretch it to weeks. Most of the time goes into cleaning and mapping, not the import itself, so start there.

What is the biggest risk in data migration?

Silent data loss: fields that are cut off, mis-mapped or not exported at all, and only noticed weeks later. A tested backup, a complete field mapping, a test migration and record-count checks after the move are the four habits that prevent it. Keeping the old system read-only for a while adds a safety net.

Should we clean data before or after migration?

Before, in the exported file. Cleaning first means you only import good records, the new system starts tidy and your counts are easy to verify. Some small fixes are easier after the move, but duplicates and format problems should be solved before import, because they multiply once data is linked.

Can we migrate data ourselves or do we need a developer?

Simple moves with built-in import tools, such as CSV into a CRM, are usually manageable in-house with this checklist. Bring in a developer when you have several linked systems, attachments, large volumes, a custom database, or when the move must be scripted through APIs to avoid downtime.

Planning a system switch?

We plan and run data migrations for small businesses: backup, mapping, test runs and verification, with no lost records. Book a free consultation.

Book a free consultation