I wrote last week about dying with dignity

Death comes for us all, and companies are no different. Eventually, the time comes when nobody wants what the company is selling. 

A dying company has two options. It can accept its fate: milk the last drops of cash from its business; give the money back to shareholders; everyone shakes hands and goes their separate ways. 

The second option is for the company to attempt to reinvent itself. The corporate equivalent of a midlife crisis. 

Corporate finance experts and investors usually don’t approve of the reinvention option. But companies do it all the same.

The impulse is understandable. No person wants to die, and no company wants to dissolve itself.

Corporate finance people usually disapprove of corporate reinvention because it rarely works. A good company is, necessarily, set in its ways. That’s how it got to be good in the first place. A good company is designed from the ground up to solve a specific problem. It succeeds and survives because its investments, reporting lines, and talent are all oriented around solving The Problem. Radically changing strategy is like trying to retrofit a boat for air travel.

I once witnessed first-hand what it’s like when a company tries to change direction. It was a mess. Our company didn’t have the skills to do the new thing. And when it hired new people in, the existing staff didn’t want to give up their power and status. We floundered for 18 months, and then abandoned the new strategy. Our CEO got the chop.

The art of the pivot

The reductio ad absurdum says all companies are doomed, and any effort to change will fail, and they should be glad their corpses will fertilise the soil.

But of course, in reality, companies change and grow and pivot. It happens all the time. The best ones are paranoid, and constantly evolving to give consumers what they want. 

One way of doing it is the Jeff Bezos trick: to avoid Amazon turning into a big inflexible bureaucracy, Bezos turned its internal functions into products that were sold on the market (like cloud computing and warehouse fulfilment). That moved middle managers closer to final customers, kept them accountable, and stopped Amazon from getting stodgy.

The nightmare decision for managers, as described by Clayton Christiansen, is when a disruptive new technology comes along that promises to undermine a company’s business model. The new technology isn’t a big deal yet. But in a couple of years it’ll probably be the standard.

It’s a nightmare for managers because adopting a disruptive new technology will necessarily mean killing off part of the existing business. It’ll mean making less money in the present, or maybe losing money for a few years. It might mean letting go of staff. 

This is hard when competitors are breathing down a company’s neck. But it’s even harder when the company is in a strong position, trying to anticipate likely future competitors, and trying to cut them off before they get strong. It’s harder precisely because there’s no pressure — who wants to make tough decisions when times are good? 

This is basically why companies rise and then fall. When they’re big, they have too much to lose by changing. Small companies with nothing to lose can build a business model around the next technology.

Bezos is a maniac who’s determined never to get complacent and never to get disrupted. When he set up the Kindle team, he told it to treat the Amazon books business (Amazon’s cash cow and most important division at the time) as a competitor to be destroyed. 

Another canonical example is Intel under Andy Grove, who read Clayton Christensen’s book when it came out in 1997 and took it to heart. He intentionally cannibalised Intel’s profits by launching a downmarket processor called the Celeron. This he did to block the lane for any upstart competitor. 

Intel has since slipped back to a distant second-place chipmaker behind the Taiwanese foundry TSMC. Why? Because Intel turned down the opportunity to make chips for the iPhone. Apple went with chips designed by ARM and manufactured by TSMC instead. The consequence is that TSMC now dominates the smartphone market, which is the biggest market for computer chips.

Why did Intel turn Apple down? It didn’t want to cannibalise its existing business by offering Apple a low price. Whereas ARM and TSMC, with nothing to lose, were happy to do it.

But even if a CEO has the right blend of paranoia, psychopathy and drive, changing course is still a very hard thing to get right. The experience of US film and TV studios shows just how hard it is. In the late 2010s, they had been making movies and TV shows for a long time, selling them to cable companies and in cinemas, and making decent money. Then Netflix came along and sold TV and movies directly to consumers. The studios got jealous and paranoid. They saw a future where they were cut out of the deal — with cable TV and cinemas replaced by direct-to-consumer models like Netflix.

Their choice was to stay with the existing TV and cinema model, or try to go head-to-head with Netflix by building their own subscription services. In going for it, they were sure to kill off their existing model. But the hope was that they’d survive for another ten years as streaming sites.

They decided to go for it. Disney, HBO, Discovery, Warner Brothers, Paramount and NBC all launched their own direct-to-consumer streaming products in the last year or so. 

You might think this was the brave, correct decision. They ruthlessly undermined their existing business to safeguard their long-run future. But the problem was, it hasn’t worked. They’re all finding it hard to bring on subscribers to their direct-to-consumer products. And in the process, they ruined their existing businesses. Now they’re stuck between two stools. 

A few weeks ago I wrote about Google, which is trying to work out right now what it should do about large language models (LLMs). LLMs look like they might be the future of internet search. But they’re unfriendly to Google’s ad-based model. It’s a terrible decision, with Google’s $1.37 trillion market cap at stake. 

The irony for Google is that it seemed to be doing the right thing. It put a lot of effort into inventing new things. Its internal AI teams are among the best in the business. They invented the technologies that made LLMs possible. 

But it might turn out that inventing god-like intelligence is the easy part. The hard part is convincing a bunch of middle managers and shareholders that the company is going to have to change.