4 ms·

This is a clever, but terrible idea. I have had over 100 different direct reports over my entire career, and in all cases, I think having this number would hav

This is a clever, but terrible idea. I have had over 100 different direct reports over my entire career, and in all cases, I think having this number would have been a counterproductive distraction. In a sense, this is how gig economy workers are managed now. They don't have a human manager, they interact with an app that evaluates them based solely on objective metrics. It's a kind of sociopathic management style. To the company, these workers are only people in the abstract sense. Practically, they are machines from which a specific output is expected ad measured. OK, let's apply this to other kinds of work. Once management (or people in general) have a single number that ranks quality, it will be used. It will be relied upon. It will be gamed. Even if they start out as honest signals, they will be corrupted. We humans are messy. We have a long and complex history of trying to rank each other in "simple, objective" ways and it almost always ends badly. The seminal work on this subject is Stephen Jay Gould's The Mismeasure of Man. I can't recommend it highly enough. It's one of the handful of books that really changed how I look at the world. Months after I finished it, new "Hey, wait a minute!" thoughts kept bubbling up at random times.

sjg007 5y ago
I think ENPS could be a lot better than what we currently have because it is short to administer and can be administered multiple times a year quickly. The only issue I could see is that the ENPS is an aggregate statistic. You actually want to tease out their marginal contribution (and I mean marginal in the statistical sense). Maybe that can be done with something like an EPS, maybe. Interesting to think about. The only issue is that this biases social skills vs technical skills. You could have a brilliant productive employee with a really low ENPS this would be someone who is really good technically but may not get good "teacher reviews". The classic example here is the research professor. These people exist in companies too and not just in academia. An academic department can be full of people who hate each other but need each other to bring in grants. Tenure is really a bet on who can bring in the grant money. Gamification does seem to be an issue though, but Performance reviews are already gamed. There is also a recency bias and the issue that business is dynamic so goals change quickly. There are attempts to resolve this but I haven't found anything decent nor do I think the scientists have. Calibration among groups also seems quite fraught too. The sales side seems to have the only realistic non-game-able metric which is the direct revenue they bring in. This stuff is not healthy regardless because we know that work has a lot of personal nuance anyway. BTW a good podcast episode on the core issue addressed from multiple fronts: https://www.nytimes.com/2022/02/25/opinion/ezra-klein-podcast-c-thi-nguyen.html https://www.nytimes.com/2022/02/25/opinion/ezra-klein-podcas... --- Or maybe having an opaque performance review process is actually the best thing because it introduces enough fog to avoid gamification and more importantly people can't get overly upset... dunno.

sjg007 5y ago

I think ENPS could be a lot better than what we currently have because it is short to administer and can be administered multiple times a year quickly. The only issue I could see is that the ENPS is an aggregate statistic. You actually want to tease out their marginal contribution (and I mean marginal in the statistical sense). Maybe that can be done with something like an EPS, maybe. Interesting to think about. The only issue is that this biases social skills vs technical skills. You could have a brilliant productive employee with a really low ENPS this would be someone who is really good technically but may not get good "teacher reviews". The classic example here is the research professor. These people exist in companies too and not just in academia. An academic department can be full of people who hate each other but need each other to bring in grants. Tenure is really a bet on who can bring in the grant money. Gamification does seem to be an issue though, but Performance reviews are already gamed. There is also a recency bias and the issue that business is dynamic so goals change quickly. There are attempts to resolve this but I haven't found anything decent nor do I think the scientists have. Calibration among groups also seems quite fraught too. The sales side seems to have the only realistic non-game-able metric which is the direct revenue they bring in. This stuff is not healthy regardless because we know that work has a lot of personal nuance anyway. BTW a good podcast episode on the core issue addressed from multiple fronts: https://www.nytimes.com/2022/02/25/opinion/ezra-klein-podcast-c-thi-nguyen.html https://www.nytimes.com/2022/02/25/opinion/ezra-klein-podcas... --- Or maybe having an opaque performance review process is actually the best thing because it introduces enough fog to avoid gamification and more importantly people can't get overly upset... dunno.