Meta terminates its low performers and does not regret it
Following cuts in 2022 and 2023, Meta targets 10% non-regrettable attrition and will terminate low performers faster.
Meta recently announced that they have a 10% non-regrettable attrition target for 2024 into 2025. 2024 provided 5% of that target, and they will achieve the remaining 5% with aggressive performance-based cuts this February. This isn’t a layoff intended to reduce headcount, since they intend to backfill the positions over the course of 2025. A memo stated…
We typically manage out people who aren't meeting expectations over the course of a year, but now we're going to do more extensive performance-based cuts during this cycle — with the intention of backfilling these roles in 2025. We won't manage out everyone who didn't meet expectations for the last period if we're optimistic about their future performance, and for those we do let go we'll provide generous severance in line with what we've provided with previous cuts.
A separate memo that was sent out later that day highlighted how this will be implemented.
Anyone who receives a rating of “Met Some” or “Did Not Meet” will be automatically added to the performance termination list.
Later in the calibration process, your Director and VP will review those with a “Met Most” rating to determine who will be terminated to meet the required 10% target. The number of people in the “Met Most” category to be terminated will vary, depending on your org’s 2024 non-regrettable attrition rate and how many people are rated “Met Some” or “Did Not Meet”
You should use the flag and notes features within the performance tool to make any recommendations about whether someone with a “Met Most” rating should be included in the performance terminations or not.
You might wonder: how should you behave if you work at a company that has a performance attrition policy? Your gut reaction might be to “work hard and get a great performance rating.” This is a reasonable answer, but it puts a lot to chance. What if you work hard on the wrong thing? What if your manager just hates your guts and just gives you a low rating? What if you don’t know how to pick between good opportunities?
So you may also find that “game the performance management criteria” is also a perfectly reasonable answer. Instead of leaving the process to chance, you change your behavior to match the way the performance criteria are written, and also make sure that everyone knows it. What does it look like to game performance management systems? Let’s look at some real-world examples.
When I started out in the industry in the 00’s, Microsoft was notorious for using a brutal stack-ranking system that they ended in 2013. This system was inspired by Jack Welsh, of GE fame. Welsh believed that the bottom 10% of his company should be fired every year to make room for higher performers.1 In Microsoft’s implementation, they were required to rate their employees on a bell curve. So if you are a manager and you are generally happy with your team, you cannot give everyone an “exceeds expectations” rating. The bell curve would Very Strongly Encourage two employees to receive a “meets expectations” rating, one would receive an “exceeds expectations” rating, and one of them at “below expectations.”
This system had lots of failure cases.
What if you want to get promoted. Why not? It comes with more money and more interesting job responsibilities. In Microsoft’s ranking system, you can make your promotion case when everyone on your team receives “meets expectations” even if they also deserve to be promoted. One approach is to join a mediocre team and excel on it. But in the absence of a mediocre team, you could simply sabotage your coworkers so that they are more likely to receive “meets expectations” ratings. Maybe you intentionally delay your responses to them, or underspecify information. Maybe you avoid pointing out pitfalls. Maybe you gossip. The sky is the limit!
Let’s say that you’ve overperformed for three years and deserve a promotion by any objective reading of the career ladder. But you work with a superstar who is obviously going to earn your team’s “exceeds expectations” every rating cycle. If you wanted to stay at Microsoft, you could sacrifice three years of working towards a promotion and find another team at Microsoft — and hope that you can somehow fight to the top of the new team. But more likely, you would start interviewing for jobs at your desired level outside of Microsoft.
Imagine that four world-class engineers are together on a team. They move heaven and earth and produce fantastic outcome after fantastic outcome. The bell curve will force one of those engineers to get a “below expectations” rating when they get their performance review. So they leave the company because they obviously have job prospects. So if you are a world-class engineer in this bell-curve stack ranking system, you really want to avoid working with other great people as much as possible. Obviously this hurts your innovative output, but it sure helps your performance rating.
If you knew that a specific employee on your team was probably going to get the team’s “below expectations” rating, then you could slack off until the next performance cycle. You don’t have to be great, you just have to be better than the guy who is destined to get the “below expectations” rating. The bell curve will force your rating to be “meets expectations.”
In fact, basically any written system can be gamed if you simply do the things in the career ladder as much as possible2. A famous promotion hack at Google was to just keep making new products. Their career ladder explicitly called out demonstrating impact through launches. In fact, it was common for promotion packets to reach the promotion committee. They would read through the packet, discuss it, and then reply, “oh yes, you definitely deserve to be at this next level. And when you launch your project and demonstrate all of that launch impact, then you will definitely get the promotion. Please apply next cycle.”
When I joined Google, it was really jarring to meet an engineer that had the following strategy:
Find an ambitious product manager.
Work with that product manager to identify an opportunity they could build in under eight months with 3-5 people, using Google’s existing infrastructure. Sometimes this was a standalone product, but often this involved adding a new service to support one (or more) already-popular products.
Launch.
Iterate for a few months, long enough to show that usage is growing.
Teach the team how to run itself.
Find another product manager and do it again.
At promotion time, just gesture towards the sea of projects he founded that were still growing.
To his credit, he worked extremely hard and aggressively networked. But many people worked very hard at Google, and he got better-than-average promotion results specifically because he also understood the promotion criteria and joined forces with people of complementary roles that also wanted to get promoted as fast as possible.
So what should you do if you are at Meta, and you would like to keep your job in the next performance cycle? Yes, of course you should work hard. But before you do all of that work, I would read every sentence of your target promotion in the career ladder and really think about who becomes a low performer, who becomes a high performer, and why?
I’m having trouble finding performance-based termination targets that Microsoft used in the 90’s and 00’s. A few unreliable sources indicate it was around 5%, so about half as severe as Jack Walsh’s GE cuts.
Ignoring the cases where the career ladder does not accurately reflect how people actually get promoted.