“I built a spreadsheet and we work off it. Except now there are 10 of us in the operation. And to use the spreadsheet, everyone has to let the others know when they’re not using it.”
That’s what the manager of a credit company said. He’s not complaining about the spreadsheet: he built it, and it solved his problem for a good while. What changed was the size of the operation. The tool that handled one person became a waiting line for ten.
In 169 meetings with 38 companies, we classified 342 pains. 161 were about spreadsheets, three times more than the second-place theme. Most people looking for a CRM aren’t switching systems. They’re leaving the spreadsheet behind. And they always get stuck on the same question: how do I do this without stopping sales halfway through?
The migration doesn’t get stuck on the system. It gets stuck on the operation.
Importing data is the easy part. What derails a migration is the month when a salesperson logs the same sale in two places. It’s the proposal that went out the old way because the customer demanded it. It’s the spreadsheet that stays open in the next tab, “just to double-check,” and within three weeks becomes the source of truth again.
A migration is a period when two tools coexist. Whoever plans that coexistence makes it through. Whoever pretends it doesn’t exist goes back to the spreadsheet and tells everyone “the CRM doesn’t work for how we operate.”
The first decision isn’t which system. It’s: how many days will the two things coexist, and who’s in charge of each one during that period.
Decide what comes along, and what stays behind for good
The most expensive mistake is wanting to bring everything: five years of history, thirty-seven columns, tabs nobody understands anymore. Split it into three piles.
Comes along. What’s alive today: active customers, deals in progress, unanswered proposals, contracts in effect. Without this, the new system starts out empty.
Comes later, if it’s missed. Closed and lost deals, customers with no activity for over a year. Useful for reporting, not for the operation.
Doesn’t come at all. Columns nobody has filled in for months, calculation tabs, free-text fields. These aren’t data: they’re symptoms. Quick test: “what decision do I make by looking at this column?” No answer within ten seconds, it doesn’t come.
The 30-day plan
Week 1: freeze and design. Nobody creates a new column or rule from today on. Map the pipeline on a sheet of paper, with no more than seven stages. Write down up to five loss reasons (“price” alone isn’t a reason). Set the required fields using the most unpopular and most effective rule: if it doesn’t change a decision you make, it’s not required.
Week 2: clean and import. Clean the “comes along” pile right in the spreadsheet: duplicates, names written three different ways, a deal closed two years ago that’s still marked open. Import fifty rows first, check them column by column, only then upload the rest. And check the number you look at most: if the spreadsheet says 84 deals totaling R$ 300 thousand, the system has to say the same. If it doesn’t match, find the difference now.
Week 3: coexist, with a clear rule. Write the rule in one sentence: “starting on day X, new deals are born in the system; the spreadsheet becomes read-only.” No “whoever prefers.” Train each person on their own deals, on screen, logging one real step forward. Fifteen minutes per person is worth more than a two-hour presentation.
Week 4: close the door. Remove write access to the spreadsheet. Run the first full month-end close through the system, even if it looks rough. That’s when the migration actually happens: when the results meeting uses the new number. Then go back to the design and cut whatever nobody used.
The four mistakes that stop the operation
Migrating and configuring at the same time. Pull the data, let the team breathe for two weeks, then automate.
Starting with automation. Automation on top of dirty data spreads errors faster. First the record becomes trustworthy, then it becomes automatic.
Making mandatory what nobody will read. Every required field is a toll. If it gives nothing back to the salesperson, they’ll fill it in with whatever on a Friday, from memory.
Giving nothing back to whoever fills it in. Show each salesperson their own stalled deals, their own unanswered proposals. The day the person opens the system on their own, the data cleans itself up.
How to know it worked
Sixty days after the switch, three questions: is the spreadsheet still being opened? Can you answer how many deals have been stalled for more than 15 days? Did the month-end close get shorter? If all three come back good, the migration took. If not, the problem is rarely the system: it’s one of the four mistakes above, and all of them are fixable.
You can do all of this on your own, and plenty of companies do. It’s worth calling for help when a sale turns into a contract, the contract turns into a work order, and the work order turns into billing: that’s operation design. DualWin has done this for more than five years, in more than 100 companies. Talk to a specialist. And if you’d rather go it alone, start with week 1: freezing the spreadsheet and mapping the pipeline on paper costs nothing and is already half the journey.
Frequently asked questions
How long does it take to migrate from a spreadsheet to a CRM?
Thirty days is a realistic starting point for a small or medium operation: one week to freeze and design, one to clean and import, one of coexistence, and one to close the spreadsheet and run the first month-end close through the system.
Do I need to bring the entire spreadsheet history into the CRM?
No. Bring what's alive: active customers, deals in progress, unanswered proposals, and contracts in effect. Closed-deal history is useful for reporting and can come in later. Columns nobody fills in shouldn't come at all.
Why does the team go back to the spreadsheet after the migration?
Almost always because the spreadsheet stayed editable, or because the new system gives nothing back to whoever fills it in. When a salesperson sees their own stalled deals in the system, they start opening it on their own.
