On Shrinking Pies

Hello again, reader. It’s been a while, I’m not going to lie. Life gets in the way, what can I say?

At any rate, I was chatting to a friend recently who was stunned by my music-listening habits (well, maybe ‘stunned’ exactly isn’t the right word; more like weirded out). I don’t have a Spotify account, and mostly listen to obscure artists on Bandcamp. I also, as the cliche calls for, collect physical media - started when I was a kid with LPs (my first one, Holy Diver by RJD, is still one of my prized possessions), and now moved to cassettes as most black metal artists tend to release on that format these days. Anyways, this whole conversation, coupled with this stat I came across, made me think of shrinking pies and bottlenecks. The association is tenuous, but bear with me.

Continuing with the example of music: if you’re an artist today, you’re not having a good time. Not only do you have to fight with horrible platforms that take your profits (like Spotify), but you also have to compete with the entirety of humanity’s musical efforts. In fact, more than 70% of music streamed nowadays is back catalogue. In other words, a strong driver for listeners is nostalgia. That’s not even addressing the looming AI song generation debacle (which I’m not even going into at the moment, as it deserves a much more thorough investigation).

There is a finite number of humans streaming music, but a critical mass of music out there. If you’re an emerging act, you’re fighting for a very small niche. Even though anyone with Internet access and a credit card can find your music, very few will, as they’re overwhelmed by choice. Music’s bottleneck is people (finite) and availability (ironically infinite). Although the pie grows, the accessible slice shrinks for artists.

Every industry has its version of this problem, and the consultancy class (mea culpa!) has spent considerable energy pretending otherwise. The adage goes: growing markets are rising tides; rising tides lift all boats; therefore, growth is good, and more is more. It’s a reassuring narrative, and like most reassuring narratives, it survives mostly because people (read: boards) need it to.

But this logic simply doesn’t hold when the cost of entry approaches zero. When digital tools, platforms, and now generative AI dissolve the barriers to participation, you don’t get opportunities distributed more evenly. You get the same opportunity contested by an exponentially larger crowd, all of whom were told - correctly, technically - that the door was open.

This is the paradox: abundance now creates scarcity - not through rarity, but through overwhelming choice. So, while the pie grows, your slice gets smaller because of abundance.

What traditional strategic thinking consistently misidentifies is where the bottleneck actually lives. The instinct is to look at the obvious constraints - capital, distribution, access to the right rooms. But in a world of infinite supply, the bottleneck moves upstream, to attention, which is stubbornly finite. There are only so many hours in a human day. There are only so many decisions a person makes about what to engage with before exhaustion sets in, and they just put on Tiësto again (Adagio¹ for Strings, anyone?).

Whoever controls the interface to attention holds the actual leverage. Spotify beat the music industry by becoming the layer between human curiosity and the catalogue - and then deciding, via algorithm, which curiosity to direct to which catalogue.

Strategy must address this. In a shrinking-slice world, it’s not just about making a better product or reaching more people. Recognising defensible positions makes the difference. An emerging act with a dedicated core of fans holds a stronger position than one chasing fleeting algorithmic virality and landing in countless passive queues.

The (trite) pie metaphor, in the end, is slightly too tidy / an oversimplification. We’re not all eating from the same pie. We’re each trying to convince people to come to our table after they’ve already eaten. Which brings us to the quite liberating conclusion: if there is no pie, perhaps the point is to stop staring at the empty dish and start baking something else. (I’ll stop with the pie metaphor now, I’m hungry.)

This is easier said than done, of course. “Just create a new market” is the kind of advice that sounds visionary/profound/unheard of in a TED Talk and evaporates the moment you try to act on it. The question that actually matters is not what to create, but how you think about it. What is your operating theory? Because without one, you’re not creating anything.

Felin, Gambardella, and Zenger (go read the paper, seriously. And thanks to Jeff for sending it over!) put it plainly: to create new value, you need a theory. Not a framework or a room full of sticky notes. A theory is a set of contrarian beliefs, seen as slightly wrong by most, but essential because frameworks organise what’s known, while theories uncover what others can’t see.

I’ll spare you the full rant about generative AI. Smarter and more patient people are making those arguments well. What interests me is the silver lining beneath it: when the internet is a flat wash of AI-generated slop - when content is infinite, surfaces are smooth, and everything reads like it was written by a confident but untroubled mind - something becomes more valuable again. Specificity. The kind of thought that could only come from a particular person, in a particular context, who believed something most people thought was wrong. In other words: theory. Actual, falsifiable, testable, stake-in-the-ground theory. The kind that says if this is true, then that follows. The kind that can be proven wrong, which is precisely what makes it useful.

The streetlight effect - or, the metaphor for the tendency to search for value only where the light already shines - is what AI optimises for, almost by definition:

A policeman sees a drunk man searching for something under a streetlight and asks what the drunk has lost. He says he lost his keys and they both look under the streetlight together. After a few minutes the policeman asks if he is sure he lost them here, and the drunk replies, no, and that he lost them in the park. The policeman asks why he is searching here, and the drunk replies, “this is where the light is”.

AI is, structurally, a very sophisticated drunk searching under the lamp. It searches where the data is, synthesises what has already been said, and produces the most plausible continuation of existing thought. It is the apotheosis of the common belief. Which means that contrarian belief - the raw material of any genuinely new theory - is exactly what it cannot generate, and exactly what becomes rarer and more valuable as the slop spreads.

So perhaps the final irony of the shrinking pie is this: the same thing that flattens everything might force us to think carefully. To stop gestating vague hunches as a strategy. To commit to theory. Here’s to hope, and now, back to Bandcamp.