You cannot cargo cult the layoffs

Set aside which company and which owner because those are divisive enough to swallow the interesting question underneath.

What’s really being asked is what happens if somebody successfully implements a massive cutting back of how a technology company operates. Assume for the sake of it that the change does work which will take anything up to a couple of years to become clear either way.

The honest answer is that almost nothing changes elsewhere, unless people can work out how it worked. That’s the constraint everybody skips. You can’t copy firing half the staff, driving out another quarter, and slashing internal services down to nothing because doing that to most companies would end them within the quarter. It appears survivable in one specific case because a lot of what was being cut was really side to the product, and shutting those things down didn’t touch what users came for.

That’s a property of that company rather than a technique. Whether you have a large layer of non-core activity that can be removed without breaking anything is a question about your company’s history and the answer for most is a lot less than the people proposing the cut believe.

Then there’s what happens afterwards which gets less attention than it deserves. Recruiting into a business that just lost three quarters of its people is hard. Anyone considering it can see what happened and has to weigh whether they’d be next, so you’re hiring from a much smaller pool at a higher price. That means the things your remaining staff can’t already do are things you may simply not be able to start and you stagnate for a while if you survive at all. Stagnation eventually alienates investors and users, in that order.

So it’s an extremely risky thing to copy. Some companies are certainly looking at it and concluding they could do the same and a few will try and some of those will fail in a way that gets much less coverage than the original attempt received.

There’s a related point worth holding onto through all of this. Most businesses don’t employ people to babysit their products. They build the product to be resilient and to run without much attention, because people whose job is watching something work are expensive and automation is cheaper. If a company really has a large number of staff doing nothing visible, that’s a fact about that company rather than an insight into how software organisations work.

Which means the lesson people want to take from this, that everybody is carrying enormous needless weight, doesn’t follow from a single example even if that example works out. It might be true somewhere. Finding out by cutting first and looking afterwards is an expensive way to run the experiment and you only get to run it once.

What makes it dangerous as a template is that the failures will be quiet. A company that cuts too deep doesn’t announce it. It gets slower, ships less, loses the people it most wanted to keep, and eighteen months later is simply less relevant than it was, with no single event anybody can point at.

That’s the outcome to weigh, and it never makes the news. And the case being copied hasn’t finished yet. Anyone drawing a lesson from it is drawing it from the first year of something that needs several years to resolve, which is roughly the same as declaring a patient cured on the way out of surgery.