Why we rebuilt Excel instead of replacing it

A vehicle leasing operation we built for was quoting entirely in Excel. Each quote was a file. A deal in negotiation would produce seven to nine of them, named the way anyone names files under pressure. Two to three hundred of these moved every month between the sales team and the approver.

Nothing was versioned. Nobody could reliably say which version a client had actually received. Two people could quote the same deal differently, and a quote could go out with outdated figures in it — the kind of mistake that either costs money quietly or costs credibility loudly when you go back to a client and ask them to accept a different number than the one they already agreed.

So: replace the spreadsheets. Obviously.

Structured form. Controlled inputs. Validation. Every field where it should be, no free typing, no way to enter a number that breaks the pricing. Clean, safe, and by every reasonable measure an improvement.

The team used it, disliked it, and went back to Excel.

They weren't building one quote. They were building three or four at once, side by side on the same sheet, with rough working in the cells around them.

Different tenures. Different residual value assumptions. Different vehicle mixes. All visible simultaneously, so they could see which structure held up. The comparison wasn't a step in their process — the comparison was the process, and it worked because everything sat next to everything else on one surface.

A form cannot do that. You can build a form that produces one quote quickly, and you can build a form that produces several quotes sequentially, but you cannot build a form where the value comes from seeing candidates adjacent to one another with scratch arithmetic in between. That's a spreadsheet. It's what spreadsheets are for.

This is the part that's hard to reach from outside a business. Ask a sales team what they need and they'll describe the quote. Watch them work and you'll see them comparing quotes. The gap between those two things is where most operations software quietly fails — it gets built, it gets rolled out, and within a month everyone is back on their own sheets with the new system used as a place to file the final answer.

A real one, inside the platform. Actual cells, actual formulas, the surface the team already knew. It took about two weeks to build. Adoption took no time, because there was nothing to adopt.

Which reintroduces the original problem: if people can type anything, they can quote anything.

The answer was to leave the surface free and put the controls at the exits.

• The formulas are locked. The team works freely in the sheet but can't change the pricing rules underneath it. Only heads can, and in practice they don't need to.
• Export is gated. A working sheet contains several candidates and a lot of rough numbers, so the system refuses to produce a document until the sheet resolves to one clear quote. Working can't accidentally become a client-facing document.
• Approval follows the path that already existed. Several quotes go to the head of operations, who picks the one that goes out — which is exactly what was happening over email, now with the versions attached to the deal rather than scattered across inboxes.

When a team refuses to use good software, the software usually isn't bad. It's usually missing something the team never thought to mention, because to them it isn't a feature — it's just how the work is done.

Excel isn't the enemy in most operations. Excel is the evidence. It shows you exactly what your system needs to be able to do, and the parts people won't give up are the parts that matter most.

The goal isn't to remove the spreadsheet. It's to put a boundary around it, so the freedom stays where it's useful and the control lands where the mistakes escape.