> Even if the fired employees are good, this keeps everyone motivated and allows you to churn through employees to find the really excellent ones worth developing. The system is not inherently good or bad, as long as all parties are aware of it.

Everything about these statements is wrong. Take it from someone who worked at Microsoft before and after they finally smartened up and ditched stack ranking: it's horrible. It creates an atmosphere of paranoia, where people sabotage each other at great cost to the company instead of working together towards common goals. Worse, contra your claim that it weeds out the worst and lets the best rise to the top, what actually happens is the best quickly get sick of the bullshit and leave, while the ones rising to the top are the ones most adept at politics and gaming the system.

Stack ranking has no redeeming qualities whatsoever, and if that's really what Tesla is doing it's enough alone to make me more bearish on them. It's that bad.

I don't know if OP meant stack ranking, but I've definitely seen approx 20% annual turnover at a well known consulting firm, by design. Either you go up or you are out. Not stack ranking, just performance review. Cut throat environment, but the upside is the firm is able to keep the very sharp ones. Also the alumni usually end up at other firms that can potentially become future clients. Seems to work very well for this consulting company.

> Seems to work very well for this consulting company.

But is it working for the rest of us?

Maybe. Maybe not. Those who chose to work for that firm know full well what they're getting into. They hope to make it. 20% of them won't. Not much different than startups?

The top three management consulting firms are ranked #1, #3, and #11 on Glassdoor's best places to work. Clearly their up-or-out policies aren't seen as horrible by their own employees. And as a consultant who works at one of those firms, I've never seen anyone sabotage someone else's performance, not once. In fact, you are probably rated best when you help support those around you. Certainly the policy causes stress, but these firms are top destinations for MBAs regardless.

https://www.glassdoor.com/Award/Best-Places-to-Work-LST_KQ0,...

People with MBAs have a very different mindset from auto workers. When you've had "this is an optimal system" drilled into your head at school by people you've been told are very smart and you should look up to, you're probably going to think it's a good system. Especially when you know that the probability of harsh financial consequences for yourself and your family from this policy is essentially zero.

The situation for auto workers is very different on both points.

These Best Places to Work awards don't really tell me if a place is actually a good place to work. It is, however, a pretty good indicator of how strong the kool-aid is at a company.

My understanding was that tech (and I consider Tesla tech even if it's inherently auto) companies were getting away from the stack rank evaluations (save a few). So I'm a bit surprised Tesla is going in that direction.

High demand workers have more opportunities and those that can (the good ones) will find a diff job that does not play this kind of churn game. So I'd thought.

There is no stack ranking involved (as far as I know, at least in SW). That's why when an engineer is evaluated as under-performing this is really about their own (manager-perceived) performance vs the expectation, and there is no question about whether they'd be in an exceptional team. Now instead of a formal PIP that we know won't get anywhere, HR isn't shy about letting go dead-weight. My impression is that the default approach from HR when a manager rate someone as under-performing is that we should let them go, unless the manager wants to setup some sort of PIP.

Occasional underperformance is normal variance as is overperformance. If you can't handle these cases you don't belong in management, and if you fire everyone for underperforming once (your suggestion) you will soon have no workers, and you will cause everyone remaining to underperform.

I think you missed the essence of my comment which was "there is no stack ranking and being evaluated as underperforming is then not a result of stack ranking".

I'm not sure what you mean by `occasional underperformance`? I'm talking about a yearly performance evaluation, in my opinion an overall underperforming for a year is not "occasional", it shows something systemic. The truth is that sometimes we make bad hiring decision (in my group it has been 2-3%). We try to work with them for a few months, set goals, but ultimately you can't always recover everyone.

A single occurrence is never systemic.

You haven't defined what you mean by "single occurrence"? To me being low one month, or during a two months project can be a single occurrence, but that wouldn't lead to a "bad perf review" (at least not in my team). Being clearly underperforming for a complete year without changes after setting up intermediate goals for improvements is not a "single occurrence". When a Sr. Staff which is less productive than other junior engineers, and requires the same level of guidance, something isn't right.

A single bad review.

> The system is not inherently good or bad, as long as all parties are aware of it.

There's a lot of problems with stack ranking, not least that it discourages cooperation and encourages direct or indirect sabotage (e.g. declining to hire new prospects who you suspect might be better than you). Wikipedia discusses many of the issues: https://www.wikiwand.com/en/Vitality_curve

There is another problem with stack ranking -- I saw this at a previous employer: some people are hired into "4s and 5s" from the onset. I heard some hiring managers upset that they would have to force rank their hand-selected elite team into 1-5 ranks, and upset at the situation, they purposefully hired weak workers with the intention of slotting them into 4 and 5 ranks. Their argument was (i dont agree):

- these weak workers wouldnt otherwise have a job

- these weak workers get an extra year of employment they "dont deserve"

- the strong team members get the 1,2,3 ranks they deserve

- the hiring manager is "forced" to do this

All that said, seems incredibly unfair to hire people directly into the chopping block without the expectation set from the start.

Wow, that sounds awful.

It's a losing situation all around, between the company not getting what it intended (firing the pre-existing dead weight, and getting rid of scapegoats) and, of course, the employees who are duped into joining just to be fired later.

Which companies do this?

It’s an oxymoron.. ‘we only hire the best and stack rank the rest’.