From Spreadsheet to CRM: A SaaS Founder's Migration Guide

Spreadsheets are great until they are terrible. For most SaaS founders, the breaking point arrives somewhere between lead 80 and lead 200. The tab count balloons. Three people are editing the same row. A deal closes and nobody updates the stage. You promise yourself you will fix the system this weekend, and then a customer calls and the system stays broken.

This guide is for the founder who has finally decided it is time. You are moving your pipeline out of Google Sheets, Airtable, or Notion and into a real CRM. The migration itself is where most founders panic, because there is an instinct to preserve everything — every column, every comment, every dead lead from 2024. Do not. The whole point of moving to a CRM is to leave roughly 30 percent of your spreadsheet bloat behind. The faster you accept that, the faster the migration goes.

What follows is the four-step playbook I have walked dozens of founders through, plus the cleanup work that needs to happen before you import a single row.

Spreadsheets feel free. They are not. The hidden cost is the time you spend reconciling tabs, rebuilding pivot tables, chasing down which version is the real one, and explaining to a new sales hire why your pipeline looks like a tax return.

There are three reasons founders delay the move longer than they should. First, sunk cost — you spent a weekend building that color-coded master sheet and you do not want to throw it away. Second, fear of disruption — a CRM migration sounds like it will break the team's rhythm right when revenue is finally moving. Third, decision fatigue — every CRM looks the same in screenshots and you cannot tell which one will actually fit.

None of those reasons survive scrutiny once you do the math. A founder spending five hours a week wrangling spreadsheets at a $200 effective hourly rate is burning $52,000 a year on coordination overhead. A real CRM costs a fraction of that and gives you back the hours.

The single biggest mistake founders make is importing dirty data into a clean CRM and expecting the CRM to fix it. It will not. Garbage in, garbage in a nicer interface. Spend a half day on cleanup before you touch the import button. You will save yourself a week of confusion later.

The cleanup has three jobs: de-duplication, normalization, and archival.

Sort the sheet by email address, then by company domain. You will find duplicates you did not know about. The same lead from two trade shows. A contact entered as Jane Smith in March and J. Smith in July. A demo signup that also bought your beta plan.

Merge them in the spreadsheet first. Pick the row with the most complete data, copy any unique fields from the duplicate row into it, then delete the duplicate. Do this in the spreadsheet, not the CRM, because spreadsheets let you bulk-edit far faster than any CRM merge UI.

Pick a single format for every field and enforce it before import.

Phone numbers: convert to E.164 format (+1 415 555 1234). Use a formula or a quick search and replace.

Stages: consolidate variants. If you have a column with values like 'demo scheduled', 'Demo Scheduled', 'demo set', and 'demo on calendar', they all need to become one canonical value.

Source attribution: turn freeform 'where did they hear about us' notes into a clean list. Linkedin, Twitter, Referral, Cold Outreach, Inbound, Trade Show. Five to seven values, no more.

Dates: make sure every date column is in the same format. ISO 8601 (YYYY-MM-DD) imports cleanly into every CRM.

Owner names: if you and a co-founder both work the pipeline, make sure rows are tagged consistently with whoever owns them. CRMs assign records to a user, not a freeform name.

This is where the 30 percent comes off. Move every lead that has not had activity in 12 months into a separate tab called Archive. Do not import them. They are not your pipeline. They are inventory you will never sell to.

The instinct will be to keep them just in case. Resist. If a CRM has 800 contacts and only 200 of them matter, every search, every report, every automation has to filter out 600 corpses. Cleanliness is leverage. The dead leads will still exist in the spreadsheet if you ever need them — you just will not be paying the cognitive tax on them every day.

Open the CRM, ignore the import button for an hour, and design your pipeline. This means deciding three things.

What are your stages? Five to seven is the right number. New, Qualified, Demo, Proposal, Negotiation, Won, Lost is a clean default for a SaaS pipeline. If you have more than seven, you are confusing stages with substages.

What are the required custom fields? Plan, MRR, contract length, billing frequency, and integration partner are common SaaS-specific fields. Build them in the CRM first, then map your spreadsheet columns onto them at import time.

What does an opportunity look like vs a contact? Most spreadsheets blur this. In a CRM, a Contact is a person, a Company is an organization, and a Deal (or Opportunity) is a specific revenue conversation. The same contact can have three deals over time. Get this distinction right at the start.

Now open your cleaned spreadsheet next to the CRM's import wizard. The wizard will ask you to map each spreadsheet column to a CRM field. Most CRMs guess correctly for obvious ones (Email to Email, Phone to Phone). The pain is in the custom fields.

Add any custom field to the CRM before you start the import. If your sheet has a column called 'plan_tier' and the CRM has no matching field yet, create it now. The import wizard will not magically conjure fields for you.

Do a small test import first. Pick five representative rows, save them as a separate CSV, and run a test import. Verify every field landed in the right place, that dates parsed correctly, and that the deal stages came through cleanly. Then delete those five test records and run the full import.

Do not import all 200 rows in one go. Split the data into three logical batches.

Batch one: active deals (anything with a real revenue conversation in the last 60 days). Maybe 30 to 80 rows.

Batch two: active leads (qualified prospects who have not yet started a real conversation). Maybe 40 to 100 rows.

Batch three: colder contacts you still want in the CRM for nurture or potential reactivation. The rest.

Import batch one. Spend a day working in the CRM with just that data — log a call, move a deal, send a quote. If something is broken (a custom field is wrong, a stage is misnamed, an automation is misfiring) you find it with 50 records, not 250. Fix it, then import batch two. Then batch three.

This takes an extra two days versus one big import. It is the difference between a smooth migration and a mess you spend a week unpicking.

Before you close the spreadsheet for good, run validation queries on the CRM and compare to the source.

Count records by stage. Does the CRM total match the spreadsheet total? It should match exactly.

Sum the deal value. Does total pipeline value in the CRM match total pipeline value in the sheet? If the numbers diverge, you have a currency formatting issue or a field mapping mistake. Fix it before retiring the source of truth.

Spot-check ten random records. Open them in both systems and confirm every field looks right.

Once validation passes, mark the spreadsheet read-only and put it in an archive folder with the date in the filename. Do not delete it for at least 90 days, in case you discover a missing field weeks later. Then forget it exists.

The spreadsheet is not your enemy. It is a museum piece now. Treat it accordingly.

Lock it (read-only sharing in Google Sheets, archive in Notion, freeze in Airtable). Rename it with a clear archive prefix and the cutover date — for example, 'Pipeline_ARCHIVE_2026-05-08'. Strip edit access from anyone who is not you. The single fastest way to corrupt a fresh CRM migration is to have a sales rep accidentally update the old sheet because their muscle memory takes them there.

Keep it for 90 days minimum. After that, if you have not opened it once, you will not need it. Archive it to cold storage and move on.

All three sources export to CSV, but the cleanup work varies.

Google Sheets: the easiest source. Export to CSV, run a sort and filter pass, you are ready. Watch out for cells with formulas that export as their computed value (usually fine) and merged cells that export weirdly (always bad — unmerge before exporting).

Airtable: you have probably been using linked records (a Deal linked to a Company linked to a Contact). CSV export flattens these into text fields, which means the relationships disappear. Plan to import Companies first, then Contacts (linking each to a Company), then Deals (linking each to both). Most CRMs let you do this in three sequential imports.

Notion databases: the messiest source. Notion's CSV export drops multi-select tags as comma-separated text, drops relations entirely, and turns rich text into stripped plaintext. Expect to do meaningful cleanup. The good news: most founders using Notion as a CRM are early enough that they only have a few hundred rows, so the cleanup is bounded.

If you are coming from a heavily customized Airtable base, budget an extra day for relationship rebuilding. If you are coming from Notion, budget an extra day for tag cleanup.

Deelo's CRM has a CSV importer that walks through field mapping, supports custom fields you have already created, and handles batch imports cleanly. Companies, Contacts, and Deals each have their own importers, so the Airtable-style 'import companies first, then contacts, then deals' workflow is the supported path.

A few specific things that help on a SaaS migration. Deelo lets you create custom fields per object before import (MRR, plan tier, contract length, integration partner) so you do not have to stuff structured data into a Notes field. Pipeline stages are configurable per pipeline, so you can model your sales motion exactly the way you describe it on a whiteboard. Deal owners assign cleanly during import via an email-to-user lookup, so you do not end up with orphaned records.

Deelo also includes Helpdesk, Marketing, and Invoicing alongside the CRM at $19 per seat per month, which matters once you start nurturing the leads you just migrated and invoicing the ones who say yes. Migration is a one-time cost. The platform you migrate into is the cost you pay every month for years.

Migration anxiety is mostly a fear of losing data that you do not actually need. Once you accept that the goal is not to preserve the spreadsheet but to leave its bad habits behind, the work gets simpler. Clean the data. Model the pipeline. Import in batches. Validate. Walk away from the old sheet.

The founders who make this move smoothly are not the ones with the cleanest spreadsheets. They are the ones who are ruthless about what they leave behind.