There's a lot of "you have to cut X off your salary budget", pick whoever you want. Then it's up to the managers to pick, very subjectively, who they want to work with the most.
FFS - I managed to dodge 4 layoffs at a startup, including my whole department being cut wholesale. My managers always kept me, even though I wouldn't call myself the "best" or "the cheapest". All of it just boils down to one thing - I'm the one that whoever manages me can rely on. Just that... being reliable.
The senior management delegates. At every moment your direct manager knows whom they are willing to part ways with if asked. Even if everyone on the team is great there is always some criteria (costs, impact on important project, institutional knowledge) that sets some people apart.
Can they? In my experience, they don't tend to do this when it comes to deciding who to lay off. I assumed this was because they can't tell who the bottom 10-15% are.
Team members can tell, but can upper management?
How do you even know that?
Usually the only way that you will see that, as a customer, is when the people who left are unable to keep up with all the firefighting and things ultimately explode. Either with outages or a security breach.
You can accumulate a _lot_ of tech debt before customers notice. For smaller companies, that's often a fatal amount of debt.
If you decide to stop paying your water bill you don't lose service that same day. Typically these layoffs either cause or are a sign of stagnation and then some scrappy startup starts eating away at your market share.
I can bet that if you bring Google Reader up, you'll get a lot more complaining - than any "horrible Google services" topic.
Never have I ever been the person with highest "overall activity, number of issues closed, merge requests completed, etc.", yet I have not been laid off in a very long time.
Last time I thought I was getting laid off, I misread the manager's comments as PIP.
How else would you measure how much work someone does, other than by how much tracked work that's attributed to them there is?
The people using these indicators probably don't care about someone who might have fewer closed issues but instead is helping out 5 people work more efficiently, but rather about people who are constantly shipping measurably less than others, stack ranking and all - in their minds, development should be viewed as some widget factory.
Just look at what happened to agile and how SAFE ended up: https://www.atlassian.com/agile/agile-at-scale
That's also how you get attempts to turn some abstract complexity evaluation (story points) into hours that could be billed to a client ahead of time, even though it probably shouldn't and doesn't work like that.
Regarding the second para, the interesting part for me is: how do these companies determine which are the unproductive devs? My impression so far is that senior management has no way of measuring this, given how little data is provided to back up decisions to RTO.
Not so sure about the third para. Why would one company be more ahead of the game than any other, if indeed LLMs are providing this supposed boost?