Proof it runs
Thirty-one weekly reports across four of my own brands, written by the system with nobody touching them, and three real problems they caught that I would have missed.
Last updated 27 August 2026.
I can't show you revenue, and I won't pretend otherwise. What I can show you is that the thing I'd build for you has been running on my own businesses, every week, for months, and that it catches real problems.
Watch it find one
Thirty-five seconds, unedited and not sped up, running against my own store. It finds that the shop doesn't link to my own Etsy listings, so anyone who lands there and likes the work can't find the other half of it. I built the thing that checks for this and it still caught me out.
The reports, by date
Four of my own brands. Every one of these is a real report that landed in my inbox on that date. Weekly cadence starts at the end of July, when the last of the four came online.
| Brand | Reports | Dates |
|---|---|---|
| ExecutorPilot | 10 | Jul 9, 14, 21, 27, 30, 31 · Aug 3, 5, 10, 17 |
| Kitchen Comfort | 9 | Jul 3, 6, 15, 20, 27 · Aug 3, 10, 17, 19 |
| Tawny | 7 | Jul 6, 15, 21, 27 · Aug 3, 10, 17 |
| PrintedPaws3D | 5 | Jul 14, 27 · Aug 3, 10, 17 |
Thirty-one reports. Nobody wrote them by hand. The consistency is the point: a check that only runs when you remember to run it isn't a check.
Three things it caught that I'd have missed
1. A sales channel that quietly stopped reporting, for four weeks. One store's sales data silently dropped out of the feed. The report didn't guess a number or carry the old one forward as if it were current. It said the data was absent, said how many weeks it had been absent, and kept saying it until it was fixed. Four weeks is how long that would have gone unnoticed if I'd been eyeballing a dashboard.
2. 1,082 of 1,094 product images with no alt text. 98.9% of one shop's images were invisible to search and to anyone using a screen reader. No dashboard anywhere shows you that number. You have to go looking, and nobody goes looking.
3. An email platform unreachable for two weeks running. The subscriber count sat there looking perfectly normal, because it was the last number successfully read. The report flagged it as fourteen days stale rather than showing it as current. That distinction is the whole job.
What one looks like
A real report, with the numbers replaced. The structure, the sourcing column and the honesty about what was and wasn't verified are exactly as delivered.
# [Brand] Growth Report — 2026-08-17 > Scheduled Mon 8/17 07:45 ET. Two access holes closed since last week, > one still open. [Channel] now has a working read path, so for the first > time in four weeks a live sales channel independently confirms the order > count instead of the scan inferring it. [Platform] is still absent — 4th > consecutive week. [ESP] unreachable for the 2nd week. Everything below is > live-verified this morning or explicitly labelled carried-forward with its > date. Nothing is estimated. | Metric | This week | Prior | Source | Verified live? | |---------------------|-----------|--------|---------------|----------------| | Orders (lifetime) | [redacted]| — | API /receipts | ✅ | | Active listings | [redacted]| [prior]| API sampler | ✅ | | Search impressions | [redacted]| [prior]| GSC API | ✅ | | Search clicks 7d | [redacted]| [prior]| GSC API | ✅ | | Revenue / sessions | not pulled| not pulled | [absent, 4th wk] | ❌ | | Images missing alt | 1,082 / 1,094 (98.9%) | 932 / 944 | per-product JSON | ✅ | | Email subscribers | [carried from 8/3] | — | [ESP] unreachable | ❌ (14 days old) |
Why the numbers are blanked and the failures aren't: the revenue figures are my business, not yours. The failures are the part that matters to you, because they're what you're buying. A report that only tells you good news is a report you should stop reading.
What this doesn't prove
It doesn't prove my shops make money. Some are small and some are still early, and I'm not going to imply otherwise to sell you something. It proves the system runs without me, tells the truth when data is missing, and finds things a person staring at a dashboard doesn't.
That's the claim. It's the only one I can back.
Common questions
Are these real reports?
Yes. Thirty-one reports from four of my own brands, written by the system on the dates listed, with the revenue figures blanked and everything else as delivered.
How often does the check run?
Every week, on a schedule, with nobody pressing a button. The weekly cadence across all four brands started at the end of July 2026.
What kinds of problems does it catch?
On my own shops: a sales channel that stopped reporting for four weeks, 1,082 of 1,094 product images with no alt text, and an email platform that was unreachable for two weeks while its numbers looked normal.
Does this prove the shops make money?
No. It proves the system runs on its own and tells the truth when data is missing. It doesn't prove revenue, and I don't claim it does.
Or run it on your own shop right now — the free checker does the outside-in part of exactly what you just watched, in about ten seconds, with no signup.