Backstory
Back in my finance days (as an intern), my boss had me read a lot of Buffett, plus a stack of other articles that shaped how I think about building, running, and investing in businesses. Between the shareholder letters and whatever economics articles I was handed, one stuck with me: Measuring the Moat by Michael J. Mauboussin (the 2016 edition).
Handy link if you also want to suffer: Measuring the Moat (2016)
The lovely folks on X love to make up new things to tout, especially now with AI being the constant hot topic. Anything for that thought leadership clout, I guess…
Recently, I keep hearing “taste is the new moat.” If you’re reading this, I’m sure you have too.
It’s my least favorite one, because it usually gets said as a rejection of AI. A sort of designer copium that our work can’t simply be replicated by an amateur with a Claude Max plan. I think it’s a genuinely dangerous belief, and kind of like pulling the wool over your own eyes.
Here’s my problem with it. A moat is a structural barrier that lasts. “Taste is the new moat” describes something that’s scarce right now. Scarcity and durability are two different properties, and the slogan treats them like the same thing.
What a moat actually is
Sources: Mauboussin’s PDF, Pat Dorsey, Warren Buffett, and other rich white guys.
Buffett’s test for a moat is pricing power. If you can raise prices and customers stay, you probably have one. If you have to pray before raising prices, you don’t. He also cared a lot about durability: will this advantage still be there in 10 or 20 years? Fast-changing industries made him nervous for exactly that reason.
Pat Dorsey, who ran equity research at Morningstar, turned Buffett’s idea into something you can actually check. His book, The Little Book That Builds Wealth (2008), covers four sources of moats, and Morningstar later added a fifth:
| Moat source | What it means | Example |
|---|---|---|
| Intangible assets | Brands, patents, licenses that keep competitors out | Coca-Cola, See’s Candies |
| Switching costs | Leaving is painful enough that customers don’t | Your bank, your ERP |
| Network effect | The product gets better as more people use it | Visa, marketplaces |
| Cost advantage | You can sell cheaper than anyone and still profit | GEICO, Costco |
| Efficient scale | The market only has room for one or two players | Pipelines, airports |
See how none of these are “we’re better at the work.”
Mauboussin’s version frames it as two questions: how much excess profit a company makes, and how long it can keep making it.
Hamilton Helmer’s 7 Powers says it in different words. A real advantage needs two things: a benefit (it helps you win) and a barrier (other people can’t copy it). Most “X is the new moat” takes focus on the first part of that sentence. But say a word enough times, it starts to lose all meaning. (Don’t even get me started on what happened to “steelman”.)
Dorsey even has a chapter called “Mistaken Moats.” Great products, great execution, and great management. Taste would go somewhere under the first two.
Why taste isn’t the moat you think it is
Taste is visible, so it’s copyable
The whole point of taste is that people can see it. Once the work ships, anyone can screenshot it, put it in a moodboard, and hand it to their designer (or their model).
We’ve watched this happen over and over. Corporate Memphis, the flat purple people with tiny heads and noodle arms, felt fresh for about five minutes. Then every startup bought the same illustration library, and it stopped saying anything about the company. Unfortunately, most people don’t believe in the fundamental principle of branding (zig when others zag) and prefer adding to the industry wallpaper.
A lot of taste is programmable
I might get a lot of hate from creatives for saying this, but here goes. A big chunk of what we call taste is best practices: spacing scales, type hierarchy, contrast ratios, not putting six fonts on one page. That stuff has been codified for years. Design linters, accessibility checkers, and design systems already catch a lot of it automatically. Now that we have system one classifiers, it’s even easier.
If you could evaluate why something was good in design or art history classes, you can turn it into guidance for an agent.
Taste is actually alpha
None of this means taste doesn’t matter and it’s not all p(doom)=0 over here.
I think a better way to see it is this: taste is alpha.
Alpha is the excess return a fund manager earns from skill. It’s real, and people pay a lot for it. But it has a few known problems:
- It lives in people. When the portfolio manager leaves, the alpha usually leaves too.
- It decays once others notice. Researchers have found that once a market anomaly gets published in an academic paper, a big chunk of the excess return disappears afterward. McLean & Pontiff, Journal of Finance, 2016.
- Eventually, yesterday’s alpha gets packaged up and sold cheaply as a “smart beta” ETF.
Taste follows the same arc. Today’s taste is tomorrow’s template.
The interesting part is that asset managers figured out how to live with this a long time ago. A fund’s real moat usually isn’t its alpha. It’s the brand, the distribution, and the clients who stay through a bad year. Alpha is what attracts the money. Something else has to keep it. It’s not so different than how I think about running Miscreants too.
So I think the job looks like this:
- Keep generating and seeking fresh alpha.
- Convert it into something durable before it decays.
- ???
- Generational wealth
Cool things to read
- Mauboussin, Measuring the Moat (2016)
- Pat Dorsey, The Little Book That Builds Wealth (2008)
- Morningstar on the five moat sources
- Hamilton Helmer, 7 Powers (2016)
- McLean & Pontiff, “Does Academic Research Destroy Stock Return Predictability?”, Journal of Finance (2016)
- Buffett’s shareholder letters