There's a spreadsheet on your computer right now that could tell you something you'd genuinely want to know, and you're not going to open it, because opening it means an hour of squinting at columns and you already know how that goes.
That's the whole problem. Not the data. The hour.
Here's the thing nobody said out loud when everyone was telling you to "learn Excel": you can skip the formulas entirely now. Paste the sheet in, ask a question the way you'd ask a person, get an answer.
No VLOOKUP. No pivot table. No YouTube tutorial about pivot tables that you close after 4 minutes.
Before anything else: take the sensitive stuff out
You're about to paste business records into a tool. 30 seconds first.
Pull out full card numbers, social security numbers, bank account numbers, and anybody's personal details that aren't yours to share. Your own revenue, your own costs, your own customer list, your own pricing, all fine. I paste those constantly.
The test: would I email this to a stranger? If no, take it out.
Do this first, it takes 2 minutes
Open whatever spreadsheet you've been avoiding. Copy the whole thing. Paste it in with this:
Here's a spreadsheet from my business.
[paste]
Before you tell me anything about it, tell me what it IS. What
is each column, what does a row represent, and what's the time
period covered?
Then tell me: what are the 3 most useful questions somebody
could answer with this data?That last part is the trick. You don't have to know what to ask. It'll tell you what's answerable, and it's usually not what you expected.
I ran this on a job list I'd looked at 100 times and it pointed out something I'd never once thought to check.
Rest of it's yours
Drop your email and the whole thing opens up right here. No waiting on a confirmation, no download, nothing to install.
You'll also get the new ones as I make them. Leave whenever, no hard feelings.
The 6 questions worth asking any business spreadsheet
Once it knows what the sheet is, these work on nearly anything. Ask them one at a time, in the same chat.
1. What's actually making the money?
Which items, services, or customers made me the most money?
Now show me the same list by profit, not revenue, if the data
supports it. If it doesn't, tell me what column I'd need to add.Revenue and profit rank differently more often than people expect. That gap is the interesting part.
2. What's quietly losing?
Is there anything here that looks like it's losing money, or
barely breaking even once I account for the time it takes?
Point me at anything suspicious and tell me why.3. What changed?
Compare the first half of this period to the second half. What
went up, what went down, and what stayed flat that you'd expect
to move?4. Who stopped?
Find anybody in here who used to be a regular and then stopped.
Give me their name, what they used to spend, and when they went
quiet.This is the one that makes people money the same week. Those aren't lost customers, they're customers nobody followed up with.
5. What's broken in the data itself?
What's wrong with this spreadsheet? Duplicates, gaps, things
entered 3 different ways, dates that don't make sense, numbers
that look like typos. List what you find.Every real spreadsheet has this. Mine had the same customer entered 4 ways and I'd been counting them as 4 people.
6. What would you check next?
Based on everything in here, what's the one thing you'd look
into next if this were your business? Be specific and tell me
why.The 3 rules that make it work
Paste the whole thing. Do not summarize it first. Do not send it "the important columns." People shrink their data down out of politeness and remove the exact information that would've answered the question. If it's too big for one paste, upload the file instead.
Include the headers. The column names are how it knows what it's looking at. A wall of numbers with no headers gets you a wall of guesses.
Argue with the first answer. It's the same rule as everything else. If something looks wrong, say so:
That number can't be right, we never did that much in March. Check it and show me your working.
It'll go back through and either find its mistake or show you why you were wrong about March. Both are useful.
When it gets a number wrong
It will, sometimes. It's reading a table, not running a calculator.
Two ways to protect yourself, and they take seconds:
Show me the rows you used to get that number.and
Do that calculation again a different way and tell me if you
get the same answer.Anything you're about to make a real decision on, spot-check one number by hand. Not all of them. One. If that one's right, the method's working.
For anything where the arithmetic genuinely has to be exact, ask it to write you the formula instead of the answer:
Don't calculate it. Give me the Excel formula that would
calculate it, and tell me where to paste it.Now the spreadsheet does the math and it's using the thing that's actually good at math.
The bigger point
Being "good with numbers" was never the skill. The skill was being willing to sit with a spreadsheet for an hour, and most people running a business don't have the hour, so the sheet just sits there.
That's what changed. Not that AI is smart about your business. It isn't, it knows nothing about your business until you tell it. It's that the hour became 3 minutes, so now you'll actually look.
Go open the one you've been avoiding. Ask it what it is. See what it says.