When a contractor should build software, and when to keep the spreadsheet
Every contractor has a spreadsheet that has become load-bearing. Sometimes it should be software. Sometimes it should stay a spreadsheet. The four tests, and what a build actually costs now.
Somewhere in your company there is a spreadsheet that, if it disappeared, would stop the business. Equipment tracking. The sub compliance list. The daily report that gets emailed at 6pm. The WIP schedule. The question is whether it should become software. Usually the answer is no. Sometimes it is a clear yes.
Four tests
More than three people type into it. One person’s spreadsheet is a tool. Four people’s spreadsheet is a system with no permissions, no history and no validation. Conflicts, overwrites and “which version is current” are the symptom.
Someone re-types data from it into something else. The daily report gets re-keyed into the job cost system. The certificate expirations get re-keyed into the email reminder. Double entry is where the errors live and where the hours go.
It needs to be used from the field. A spreadsheet on a phone on a jobsite is not a thing that works. If the superintendent needs to enter it or read it, it needs to be an app.
The owner asks for a number it cannot give. “What is our backlog by crew for the next six weeks?” If the answer requires an afternoon of pivot tables, the spreadsheet has outgrown itself.
Two of four and it is worth a conversation. Three and it is worth a discovery sprint.
What a build costs now
Five years ago, a custom application for a twenty-person contractor was six figures and a year. The stack we build on (Convex for the data and backend, Next.js for the screens, Clerk for login, Netlify for hosting) changed that. An MVP, meaning login, the data model, the core screens and one or two integrations, is $24,500 to $45,000 and live in 30 days. It costs a few dollars a month to run.
When to keep the spreadsheet
- One person uses it and they are good at it.
- It changes shape every month because the business is still figuring out the process. Build software for a process, not for a search for one.
- The pain is real but under an hour a week.
When to build
- The spreadsheet is the system of record for something that matters (compliance, safety, money).
- The field needs it.
- You are hiring someone whose job would mostly be maintaining the spreadsheet.
How to find out
A one-week discovery sprint maps the workflow with the people who do it, mocks the screens with your real data, and produces a build plan with a fixed price. If the answer is “keep the spreadsheet,” you will hear that too, and the $3,500 bought you a documented process.
Fifteen minutes. No pitch deck.
Tell me what you build and where the leads come from today. I will tell you what I would change, and what it would cost. If it is not a fit, I will say so on the call.