Or reach editorial directly: marginalia-j28t75@polsia.app
From the latest issue
What a paid research issue actually looks like.
Two picks per issue, each carrying documented demand proof, unit economics back-solved from $1M, comparable exits, a Confidence Margin score, and the operational assumptions you have to test.
From the latest issue
Explore the rest
Pricing, the free preview, every pick in full, and how we differ from the usual listicles.
See what a paid subscription covers, browse the free preview we send weekly, and read the head-to-head against low-cost “30-ideas-a-week” newsletters.
How it works
How a raw signal becomes a Confidence Margin — in three steps.
Each pick in the Friday packet runs the same pipeline below: scout paid demand, work the math back to a $1M floor, then score honestly with the Confidence Margin — a band, never a 0–100 number.
- 01
Demand signals
We scout paid demand signals weekly.
Search volume, marketplace listing counts, and pricing pages — each cited with source and retrieval date. Adjacent exits corroborate; they are never the signal.
- 02
Unit economics
We work unit economics backward from $1M.
Solve price × volume × margin so annual run rate closes at $1M. The anchor rows are shown, never folded into a summary number.
- 03
Confidence Margin
We score every pick with a published Confidence Margin.
A band, not a 0–100 number. Two or more independent signals and clean unit math push it high; a load-bearing risk that can’t be hedged holds it down.
Pick the path that fits — the free weekly preview, or the full Friday packet.
Voices
Sourced research, kept honest by readers.
Three signals readers say they came back for: numbers, citations, the recap.
- 1,800+
Paid subscribers · weekly preview open to all
The Confidence Margin is the only reason I trust a pick before I build it.
As cited in
- · The Marginal Investor
- · Indie Hackers digest
- · Not Boring by Packy McCormick (issue ref)