Hey,
Quick update from last week’s email. I told you Tuesday (May 19) that I had shipped the GUM Toolkit. Technically I had not. I had built it and put it on Gumroad as “Unpublished” — the product page existed but it said “Not currently for sale” if you clicked the link.
I just published the actual product. Three days late.
What was live before vs. now
Before today (May 22): Medium article live → click Gumroad link → product page → “Not for sale” → bounce. LinkedIn post posted then deleted Tuesday afternoon. Beehiiv newsletter (you) sent → click link → same dead end. dev.to articles footer link → same dead end.
For seventy-two hours, every single traffic-generating channel was pointing buyers at a closed storefront.
Now:
English toolkit ($19): https://kyb8801.gumroad.com/l/gum-toolkit
- Korean toolkit v1.1 (₩30,000): https://kyb8801.gumroad.com/l/gum-toolkit-kr
Both with cover image, full description, KOLAS audit checklist, reference PDF, README. 30-day money-back. Custom thank-you receipt routing questions to my inbox.
Honest about the 3-day delay
The five-minute upload step was blocked for three days. Not by Gumroad. By me. Best guess at the brake: I told you Tuesday the toolkit “is for ten thousand people, not a billion — and that is the entire reason it might work.” Writing that sentence and clicking publish are different actions. The first commits to the audience. The second commits to the experiment that might prove me wrong. I caught myself doing the same thing in audit reports five years ago. Engineers would not file the finding for two weeks. The data was sitting on their hard drive. The finding was unambiguous. They were not lazy. They were avoiding the moment after filing.
The fourteen days that matter
The verdict on whether this niche works gets returned on June 5.
The metric is binary:
>= 1 sale → niche hypothesis holds → build next product (ISO 17025 Certificate Generator, $99, already at alpha)
- 0 sales → niche fails on cold marketplace discovery → pivot to direct outreach (KOLAS / KTR / KRISS / A2LA lab QA managers via LinkedIn DM, plan ready)
I will email you on June 5 either way. No spin.
What you can do (if you want)
Three things, descending usefulness:
Forward this email to one calibration / metrology / QA engineer you know. If they reply with “I would have paid $19 for this” — you have just done a $5 favor that costs 10 seconds.
2. Tell me what you would buy next for $19-$99 in this niche. Cert generator is queued; what else?
3. Ignore me until June 5. Honestly fine. The data will be the next email regardless.
Onward.
— YB
Toolkit (EN): kyb8801.gumroad.com/l/gum-toolkit
Toolkit (KR): kyb8801.gumroad.com/l/gum-toolkit-kr
Repo: github.com/kyb8801/measurement-uncertainty-mcp