ColurianCOLURIAN
← All articles

How to move from spreadsheets to a CRM without losing anything

A step-by-step plan for moving customer data out of spreadsheets and into a CRM: what to clean first, what order to import in, and how to handle the rows that fail.

Colurian team · · 4 min read
switchinghow-to

Most small businesses run on spreadsheets longer than they should. One sheet for leads, another for customers, a tab per salesperson, a column called "notes 2". It works until two people edit the same file, a follow-up gets missed, or someone asks a question the sheet can't answer, like "which customers haven't heard from us in 90 days?"

Moving to a CRM fixes that, but only if the data arrives in good shape. This guide covers how to plan the move, clean the data, import it in the right order, and deal with the rows that don't go in on the first try.

Before you export anything

Decide what each spreadsheet actually is

Spreadsheets mix things together. A CRM keeps them apart. Go through your sheets and sort every list into one of these:

  • Accounts: the companies or households you do business with
  • Contacts: the people at those accounts
  • Leads: people or companies you haven't qualified yet
  • Opportunities: deals, with a value, stage and expected close date
  • Activities: calls, meetings and to-dos
  • Anything specific to your business: matters, engagements, placements, donations, orders

If one sheet has a company name, a contact name and a deal amount on every row, it's really three lists. Split it before you import.

Pick what to leave behind

A move is the best chance you'll get to drop data nobody uses. Leads from five years ago who never answered, duplicate test rows, columns nobody has filled in since 2021. If you aren't sure, keep the old spreadsheet as an archive rather than importing everything.

Clean the data first

Cleaning in a spreadsheet is faster than cleaning in a CRM. Spend an hour on these:

  1. One header row. Remove merged cells, title rows and totals at the bottom.
  2. One value per cell. Split "Jane Smith" into first and last name if your CRM wants both. Split "phone / mobile" into two columns.
  3. Consistent values. "CA", "Calif." and "California" should be one thing. The same goes for status columns like "Hot", "hot" and "HOT!!".
  4. Real dates. Make sure date columns are dates, not text, and use one format throughout.
  5. Duplicates. Sort by email and by company name, and merge the obvious repeats.
  6. An ID column. If you have more than one sheet that refers to the same company, give each company a stable ID or exact name, so contacts and deals can be matched to it later.

Import in the right order

Records depend on each other. A contact belongs to an account, and an opportunity belongs to an account too. Import the parents first:

  1. Accounts
  2. Contacts, matched to their account
  3. Opportunities, matched to their account
  4. Leads, which stand on their own
  5. Activities and anything else

If you import contacts before their accounts exist, they arrive without a company and you'll have to link them by hand.

Start with a test file

Don't import 4,000 rows on the first try. Take 20 rows from each file and import those first. Then check:

  • Did every column land in the right field?
  • Do dates, currencies and phone numbers look right?
  • Did picklist values like status and stage come through, or were they rejected?
  • Are contacts linked to the right accounts?

Fix the mapping or the file, delete the test records, and then run the full import.

Expect some rows to fail

On a real import, some rows will fail: a malformed email address, a stage name that doesn't exist, a required field left blank. That's normal. What matters is how you handle it:

  • Get the failed rows back as a file, with the reason next to each one.
  • Fix only those rows and import that file, instead of re-running the whole thing and creating duplicates.
  • Group failures by reason. Two hundred rows failing for the same reason is usually one fix.

The first week after the move

The data is in. Now make sure people use the CRM instead of drifting back to the spreadsheet:

  • Make the old sheets read-only, so there's one place to update.
  • Give each person a list view of their own open leads and deals.
  • Set up one or two automations that save time right away, like a task for every new lead.
  • Build one report people want, such as the pipeline by stage, and review it in your weekly meeting.

How Colurian handles it

Colurian's importer takes .csv and .xlsx files, one object at a time. Columns are matched to fields automatically, you can change any match and preview the first row before importing, and you can download a template for any object if you'd rather paste into a ready-made sheet.

Rows that duplicate an existing record are skipped, including duplicates within the same file. Failed rows are grouped by reason, and you can download just those rows with an Error column, fix them, and import that file. If an upload is interrupted, it picks up from the row where it stopped. It works the same way for standard objects and for custom objects, including an industry edition's objects like matters or engagements.

Most teams move themselves in an afternoon. For a messy legacy export, custom objects or a complicated permission model, we also do the setup as a paid project.

Book a 30-minute demo and bring a sample export. We'll import it live on the call.