Angus Duncan · 31 July 2026 Written for Molly

I killed my own idea this morning.

You were going to open this and find a pitch for a peptide testing lab. I'd already bought the domain. Then I did the research, and the research said no.

What happened

The idea was an independent lab that tells you what's really in the vial.

Peptides are a big, fast, badly-policed market. People inject compounds bought from websites that publish certificates nobody checks. My plan was to be the lab that checks: test everything, publish the results, and make our seal mean the product is clean.

Before building anything I spent a day mapping where the money in this market actually sits. Three things came back that ended it.

01 · The lab already exists

At least eight labs already do this. Several are in the US, several hold ISO 17025 accreditation, and the market leader, Janoshik in Prague, has ten years of published results and no accreditation at all. Reputation is what buyers trust in this market, and it takes years to build. I'd be the ninth.

02 · Demand for testing is falling

As peptides went mainstream, spend on verification per buyer dropped 88%, to about $8. The whole independent testing market runs about $6M a year. Entry costs $120,000–$200,000 and 12–24 months to accreditation.

03 · The customers are being prosecuted

Over 50 FDA warning letters since September 2025. Peptide Sciences, doing $7M a month, shut down last October. Paradigm Peptides' operators were criminally charged in February. "Research use only" has been explicitly rejected as a legal defence. I'd be selling picks and shovels to a gold rush the government is closing down.

So there's no lab. There's a map of this market instead, with every number sourced, including the ones that killed the idea. At the end of it is a different business that needs no lab, no equipment and no capital.

Read the market map 11 sections · sourced

Why I'm showing you this instead

The research cost me one day. The lab would have cost two years and $150,000.

I think that's the thing worth knowing about me, and it's harder to demonstrate than enthusiasm. I already had the domain. Building the pitch deck would have been the easy, satisfying move. I did the arithmetic first and then argued myself out of my own idea.

I'm 15. I taught myself to build. Here's what the last fourteen months look like:

Output

112 75 JUN 2025 JUL 2026 learning
Repositories created per week · github.com/hangryclaude · pulled from the GitHub API, 31 July 2026

You should see the unflattering half of that chart too, because you'd find it anyway:

About a third are stubs. I try a lot of things and most of them don't survive contact with the details. The peptide lab is the most recent one, and it died faster than usual because I checked before I built.

The ask

I'm not asking you for money.

What I want is someone who has run a business reading my reasoning and telling me where it's wrong. That's the one thing I can't get by working harder.

Three things, concretely:

If you read the map and think the whole thing is wrong, that's a useful afternoon for me too.