There's a spreadsheet you've been avoiding. The job-costing sheet with a formula nobody remembers building. The bid template that breaks every time you copy a row. The hours export that takes half a morning to clean before it's any use. You keep meaning to fix it. You won't — because fixing spreadsheets is nobody's favorite afternoon, and the person who built the last one doesn't work here anymore.

Here's what took me too long to believe: you don't have to. You don't need to learn another Excel function, sit through another tutorial, or wait on the one person in the office who actually understands pivot tables. You describe what you want in plain English, and AI builds it or fixes it.

This is the same move we keep coming back to: you don't master the software button-by-button, you tell it what "done" looks like and let it work out the how. "I've got a list of jobs with costs and dates — give me total cost by job type for the quarter, and flag anything that ran over its estimate." That's not a formula. That's a sentence. The AI writes the formula.

That flips who has to be the expert. You're the one who knows what the numbers mean, which jobs were ugly, and whether a total looks off. The AI is the one who remembers how SUMIFS works so you don't have to. You bring the judgment; it brings the syntax. That's a good trade.

Say you've got a messy export from your field software — one row per work order, with columns for job, tech, hours, material cost, and invoice total. You want to know which kinds of jobs actually make money, because your gut says the small service calls are quietly bleeding you.

Old way: a couple of hours building a pivot table, fighting the date formats, second-guessing the formulas, giving up halfway. New way: paste it in (or upload the file) and say, "Group these by job type. For each, show total revenue, total cost, margin, and average margin per job. Sort by total margin, and tell me which job types are below 20%." A minute later you've got the answer you've been meaning to dig up for a year — and you can keep going: "Now show me just the service calls under $500 and their average margin." That's a conversation, not a spreadsheet project. And the answer might change what you bid next month.

AI will hand you a confident, tidy spreadsheet that is sometimes confidently wrong — a mis-set range, a row left out, the wrong column summed. Treat it like a sharp new hire's first draft: useful, not trusted yet.

So before anything leaves your hands, spot-check it. Pick one row you already know the answer to and make sure the sheet agrees. Eyeball the grand total against your gut — if it feels too high or too low, it probably is. And never send a bid or an invoice off a number you haven't verified — "the spreadsheet said so" is not a sentence you want to say to a customer. Two minutes of checking keeps the time you just saved from turning into a mess.

Cleaning and summarizing a sheet in the moment is rung one. Once your business is written down — what your job types are, what "good margin" means to you, where the data lives — you can hand off the whole chore: an assistant that takes the weekly export, builds the same job-cost report the same way every Monday, and flags the jobs that slipped before you've finished your coffee. The spreadsheet you've been avoiding turns into a report that runs itself. We'll get there in a future issue — but it starts with the small version above.

You don't have to learn the software. You have to describe the result and check the work. That's a skill you already have — it's how you run a crew.

Hit reply and tell me: what's the one spreadsheet you've been avoiding? I read everything — and it tells me what to cover next.

Next week, something a little different: instead of a how-to, I'll start running the latest AI headlines through the only filter that matters — does this actually change anything for a shop like yours?