Meta layoffs

Mar 19, 2023

Last week Meta announced another 10,000 upcoming layoffs.

Layoffs are terrible, of course, but a couple points I found interesting. Firstly, they’re explicitly flattening the company’s structure:

In our Year of Efficiency, we will make our organization flatter by removing multiple layers of management. As part of this, we will ask many managers to become individual contributors. We’ll also have individual contributors report into almost every level — not just the bottom — so information flow between people doing the work and management will be faster.

The last ten years have seen the rise of the cult of engineering management, with many tech companies treating engineering management as a more important facility than product/engineering. When I came to the valley in 2012, I had a manager who wrote as much code and shipped as much as I did, and prevailing wisdom was that this was a good thing. We’d hear that Google would purposely keep high IC/manager ratios to keep overhead down and make micromanagement impossible. Heavy hitters like GitHub even had an explicit “no managers” policy that they bragged about at conferences.

By the end of my tenure at Stripe in 2021 this had done a total 180. I had managers who by design had no function in the R&D org at all – they didn’t do product work, they didn’t do operations or hold a pager, they didn’t help with incoming cross-team asks, they didn’t manage the JIRA board, and a lot of the time they didn’t even run our meetings (deferring that to ICs); and the prevailing wisdom became that this was a good thing. A manager writing code would have been an absolutely jaw-dropping ludicrous idea – their function was to engage in high-level planning, manage people by doing 1:1s and participating in the stack ranking process, and they had four layers above them doing the same thing.

I don’t have anything against managers, many of whom are lovely people, but I’ve never seen any evidence that the manager-heavy / single-function-manager org is more productive than the alternative. Any non-trivial decisions had to be sent up and down the chain of command, committing to any decision was difficult because there were so many people involved, more documents were written than code, and communication and meeting overheads were extreme. 1

But every tech org was copying each others practices, and the familiar top-heavy pyramid was everywhere. I’d lost hope that there was any other model for organizing a company. But an organization as large as Meta moving away from it is a huge development, and this of course shortly after similar (and more controversial) decisions being made at Twitter. Is it too much to hope that the last decade’s intense infatuation with middle management was a ZIRP phenomenon?

And now, since I didn’t piss enough people off yet, that brings me to another point from Zuckerberg:

We’re committed to distributed work. That means we’re also committed to continuously refining our model to make this work as effectively as possible.

Our early analysis of performance data suggests that engineers who either joined Meta in-person and then transferred to remote or remained in-person performed better on average than people who joined remotely. This analysis also shows that engineers earlier in their career perform better on average when they work in-person with teammates at least three days a week. This requires further study, but our hypothesis is that it is still easier to build trust in person and that those relationships help us work more effectively.

Return to office is a deeply contentious issue right now. The overwhelming opinion amongst developer types is that remote is better in every way, and only idiots would try to make them return to office, and if those idiots dared try, it’d backfire because all their workers would immediately quit and go somewhere else. And yet, we see almost every high-functioning organization encouraging their employees to come back, with big ones like Amazon, Apple, and Google starting to require three days a week in office.

The developers’ answer to that would be that this is nothing more than an obsession with pre-2020 times driven by malign dogma on the part of the evil ghouls at the top of the corporate ladder. But as we see, companies like Meta have been making an effort to quantify the effect of remote work on productivity, and concluded not only that there is some, but that it’s significant enough to warrant encouraging an unpopular return.

To me, although I’d certainly concede that WFH pares down meeting culture and certainly has benefits in logistics, it’s always been fairly obvious that in-person collaboration is more effective, produces a much stronger culture, and I suspect is important especially to larger orgs because it helps to even out performance. High caliber, experienced employees probably get just as much done at home at the office, or maybe more, but junior employees with less direction struggle more, and WFH provides the ultimate cover for less motivated people (which in a larger company will always be a percentage of the total) that might otherwise be held more accountable in-person. (Obviously, some speculation here on my part.)

1 By contrast, where I work today we have managers, but in relative moderation. By my measure, this model seems to work pretty well.

Did I make a mistake? Please consider sending a pull request.