The Subprime Audience Test
If you wouldn't consume your own content, don't publish it
- Difficulty
- Easy
- Time to result
- ~ongoing to results
- Steps
- 5
- Confidence
- 68%
A subprime audience is the exit liquidity of audience capture. Like the junk mortgages of 2008, it is built from content the creator themselves wouldn't consume, so from the outside it looks like a portfolio of A-grade fans but is actually full of C-grade attention. The core test is a single question: would you consume your own content? If not, don't post it, because you have no product-market fit even with yourself. The method then swaps the metric, optimizing for depth (who bookmarks it, who rereads it ten times, who it genuinely changes) over width (raw likes and views). Finally it bounds experimentation: set brand guardrails first, because pure A/B testing with no guardrails always drifts toward the lowest common denominator.
Origin
George Mack coined 'subprime audience' by analogy with 2008's subprime mortgages, pairing it with the content razor, would you consume your own content, and the 'golden rule of the internet' that unbounded A/B testing always ends at a porn site.
Core principles
- 01A subprime audience is one you built with content you don't even like
- 02The tell of a subprime mortgage, and a subprime audience, is the seller wouldn't buy it
- 03Optimise for depth, not width: who reads it ten times beats how many liked it once
- 04Set guardrails first, then A/B test inside them — pure A/B testing ends at a porn site
How to run it
- 1
Apply the consumption test
Before publishing, ask whether you would genuinely consume this piece yourself.
Pro tip A good tell of a grift: the seller wouldn't use or buy the thing unpaid.
- 2
Kill content you wouldn't own
If the honest answer is no, don't post it, however well it might perform.
Watch out Content you don't respect can win the objective metrics and still poison your audience.
- 3
Switch to depth metrics
Track bookmarks, rereads and who shares it, not raw likes and views.
Pro tip A thousand likes from 100 people who reread it beats a million fleeting likes.
- 4
Set brand guardrails
Define what you will and won't make before you start optimizing anything.
- 5
A/B test inside the lines
Experiment only within your guardrails so you don't drift to the lowest common denominator.
Watch out Pure science with no guardrails ends at a porn site; pure art with no testing ends ineffective.
In the wild
Chris found that almost every single person who came to his live shows was someone he'd happily go for a coffee with, into cool stuff and working on interesting things. That reassured him the content he actually cared about, rather than trending gossip, was attracting exactly the audience he wanted.
→ Confirmed that depth-first content builds a high-quality, like-minded audience.
Making trending gossip content because Diddy videos get two million plays wins short-term views, but compounds into an audience that's there for Diddy, then Paris Hilton, then Kardashian gossip. Keep winning that game and you end up surrounded by people who are nothing like you and don't like what you actually like.
→ Objective-metric wins that leave you captured by an audience unlike yourself.
Common mistakes
Optimising for likes over depth
Doubling down on the million-like post while ignoring the thousand-like post that changed 200 lives.
A/B testing with no guardrails
Letting tests decide everything drifts your brand toward red-meat content and audience capture.
Publishing what you'd never consume
Making content you don't respect guarantees, at best, an audience unlike you and, at worst, a market of zero.
Is it for you?
Best for
Creators, writers and founders building a durable audience they actually want to serve.
Not ideal for
Pure performance-marketing plays where the goal is reach for its own sake, not resonance.
From the transcript
“a good indicator of a subprime audience... is the person creating the content wouldn't consume it”
“would you consume your own content if not don't post it”
“if you run enough AB tests you'll eventually end up with a porn site”
From the episode
13 Life-Changing Ideas You’ve Never Heard Of - George Mack - #801
George Mack