🔒 This is a free preview of a subscribers-only post
Welcome to part 2 of my series on disruption theory!
In part 1, I laid out the classic version of the theory, as Clay Christensen articulated it in the 90s. So now we’re ready to meet the challengers.
First up, of course, is Ben Thompson. He’s the author of Stratechery, creator of “Aggregation Theory”, and the undisputed GOAT of strategy newslettering.
To be clear, Ben Thompson is a huge fan of Clay Christensen. He’s said he thinks disruption theory is “95% right” — but the other 5% must be pretty important to him, given that he’s been developing a critique of it ever since he was a MBA student back in 2010; three years before he founded Stratechery.
Here’s a timeline of this train of thought:
Obsoletive, 2013
Best, 2014
Beyond Disruption, 2015
There’s a lot of material here, but thankfully it all revolves around one really simple idea:
What if some things can never be “too good”?
1
A core pillar of disruption theory is the belief that technology tends to improve past what people need. Whenever this “too good” circumstance happens, an opportunity emerges for disruptors to gain a foothold in the low end of the market with a different kind of product. Specifically, a modularly assembled product, rather than an integrated one. Let me explain what that means.
In the early days of a new product category, the pioneers tend to be highly integrated — meaning, they prefer to develop lots of stuff in-house. This happens because they need a lot of control in order to figure out how a new thing should work. The canonical example of this is IBM, which when they originally developed the computer did literally everything in-house.
But over time, the architecture stabilized and it became possible for specialists to enter the market. There were companies that just made software, and other companies that just made hard drives, and CPUs, and memory chips, etc. The system “disintegrated” — it broke down from an integrated architecture into a modular one.
This happens because modularization has three core advantages, according to Christensen. When it was just up to IBM to make the software, users had extremely limited options. But once anyone could make software, users had lots of new choices! The same pattern of increased choice and personalization was replicated in other parts of the system, so customers could tailor the machine to suit their needs more easily — all they had to do was swap out a component.
There’s a second benefit to modularization, too: increased speed of development. When there’s a market of several companies all competing to ship the next generation of CPUs, you often get faster progress than if one company is trying to do everything.
A third benefit is decreased cost. Usually companies are more efficient when they’re narrowly focused, and have less overhead costs.
But there is one big drawback: modular systems don’t always work. Nobody controls the whole system, so they’re quite prone to jankiness.
A great example of this is Windows in the 90s. Microsoft made the operating system, but they didn’t control the computers it got installed on. So when computer makers decided to load a bunch of stuff (affectionately known as “craplets”) onto their machines before selling them to customers, they couldn’t do anything about it.
But consumers put up with it, because of the power of the modular architecture. Windows worked well enough, it had a huge selection of apps that ran on it, and every hardware maker supported it. All the best hardware components (like Intel) primarily were designed around it. So you had fast, cheap machines with tons of programs made for it. A little pain around the edges was better than living the painful and isolated life of a Macintosh user.
2
Clay Christensen says that when performance of the integrated system is “good enough” that’s when the modular competitors have an opening.
But what if “good enough” never happens for many customers?
This is what Ben Thompson says is the case in product categories like smartphones, clothing, cars, and gaming consoles. The thing these all have in common is they’re bought by individual humans, not businesses. The user is the buyer, and so the decision is based on what people feel they want, not what a checklist of features says they need.
And what people really want — infinitely want, Thompson says — is a good user experience.
Subscribe to the Everything bundle to keep reading this post!
Subscribers also get full access to Superorganizers, Praxis, Napkin Math, and of course, the full Divinations archives. This series on disruption theory pairs particularly well with Clay Christensen’s work on modularity, which I covered in Finding Power.
Can’t afford a subscription for any reason? Email me and we’ll work something out!