Revenue tells you what sold. Profit tells you what to scale.
Two free tools that read your Shopify or Amazon exports and show you the difference, product by product.
Most of us run the business off the sales report. It updates itself, and it feels like the truth.
It's true, as far as it goes. What it can't show you is what each sale cost to make and to sell. Put landed cost, platform fees and repeat behavior next to that revenue and the order changes. Products you barely think about are carrying the business. Products you protect are being carried. (Here's what that looked like in one of mine.)
That second list is the one worth scaling on.
In my own brands I keep it running in a data warehouse, so it's current every Monday. If you're earlier than that, I built these two skills so you can see the same thing on your own exports this afternoon.
Free, no email, no signup. Whatever they find is yours.
Shopify Profit Analysis
What it reads
Your orders export, your products export, and your payouts export if you have it.
What it gives back
- Contribution margin per product: what you keep after product cost and card fees
- A role map showing which products are flywheels, which are bridges, and which are treadmills
- Twelve-month customer value by first purchase, so you know which products bring in the customers worth having
- What your below-average products left on the table, in dollars
Amazon Profit Analysis
What it reads
Your transaction report, All Listings, Business Reports, and Brand Analytics if you are brand-registered.
What it gives back
- Contribution per listing, using the actual fees in your settlement file, not a calculator estimate
- Contribution per session: what one shopper landing on each listing is worth to you
- Velocity from your settlement dates, so you can see what's accelerating into Q4 and what's dying off
- A fee ledger that flags every listing where weight or size tier is eating the margin
How to run one
Each file is a Claude skill. You hand it over once, and from then on Claude knows how to run this particular analysis properly. You don't need to write anything or install software.
- 1
Download the file above and add it to Claude as a skill.
One upload. You only do this once.
- 2
Export your data and upload it.
The checklist below tells you which screens to pull from. Budget about twenty minutes.
- 3
Say "run the profit analysis."
It'll ask about your product costs, then hand back a report, a workbook, and a cost map you can reuse next time.
What to export
The tools give you click-by-click paths as you go, so you don't need to get this right in advance. This is just so you know what you're in for.
Shopify
- Orders → Export → all orders, plain CSV. Twenty-four months minimum, all history is better.
- Products → Export → all products, plain CSV.
- Finances → Payouts → Export. Optional, but it replaces an estimated fee rate with your real one.
Amazon Seller Central
- Payments → Reports Repository → Custom Unified Transaction report. Zip it before uploading.
- Listing Reports → All Listings Report.
- Business Reports → Detail Page Sales and Traffic by Child Item, same date range.
- Brand Analytics → Repeat Purchase Behavior, ASIN view. Brand-registered sellers only.
You don't need perfect cost data
Landed cost per SKU is the one input nobody has lying around in the right format. It's where most people stop.
If you have it, the analysis tells you what you actually earn. If you have part of it, it uses what you have and flags the rest.
And if you have none of it, the tool still runs. You'll get your repeat behavior, your seasonality, your fee picture, and on Shopify, which products bring in your most valuable customers. What's missing is the profit question itself, and the report says so plainly instead of leaving it out.
Run it with what you have today. A rough supplier price per SKU is enough to start.
What happens to your data
It stays with you.
These tools run inside your own Claude account, on files you upload there. I never see them, and there's no step where anything gets sent to me.
Inside the analysis, customer emails are hashed the moment they're read. That's needed only to work out who bought twice. Every name, address and phone column gets dropped before anything is computed, so no customer detail reaches any output.
If this is useful, the writing probably is too
These came out of running my own brands. Same analysis, same arguments about which products deserve the next dollar.
Every two or three weeks I write up what I'm doing: what worked, what didn't, and the numbers underneath it.
You'll get the letter. That's it.
Read the newsletterWhat it looks like when it never goes stale
A report is a photograph. It's true the day you run it and it starts aging that afternoon. Your next cost change, your next fee tier, your next supplier increase, and the picture stops matching the business.
So picture the same analysis arriving every Monday, already current.
You open your inbox and the week is already read for you. What you sold. What you kept, by brand and by channel and by SKU. Which products moved outside their normal range and deserve ten minutes today. Your marketing spend and your inventory sit inside the picture instead of missing from it, so the margins are honest and the restock flags arrive before they're urgent. Nobody ran anything to make that happen. Nothing was exported.
Six months of that and you stop asking which products deserve to scale, because you already know. The questions get better instead: what to launch next, what to stop making, what a supplier increase does to your December, what you'd need to believe to buy the brand down the road.
That's an Ecom Data Foundation. The same models installed on your live data, running in accounts you own, refreshing themselves. It's the first thing I build in any brand I take over, because I won't operate without it.
Or have me run it
People write to me after this for two reasons. Either they'd rather hand over the exports than work through them, or a number came back they weren't expecting and they want to talk it through.
There's a third, and it's the most common. Pulling product costs together is itself the problem. That's what happens when a business grows faster than its bookkeeping, and it's the thing I fix first in every brand I take over.
A couple of honest sentences about your situation beats a polished pitch. I read everything myself.
Let's talk