Comment by revvx | Hacker News Reader

Comment by revvx | Hacker News Reader

Parent Full thread

That might sound like the most obvious thing ever written here to the point of sounding stupid, but:

In my experience as a manager, ANY kind of micro-level timekeeping or micromanagement is not only dehumanizing, but also an utterly unsustainable waste of time and resources and a recipe to let deadlines slip by.

How much time someone spends in front of their computer, how focused they are, and how their process works only gives bullshit metrics and teaches bad employees to game the system. By "squeezing productivity" you teach underachievers how to fake work while still underdelivering, while overachievers will get burned out quickly.

You'll fool yourself into believing that everything is "in control" but productivity will lower in a boiling frog effect that you'll only notice after several months of having everyone glued to their screens.

What I learned after failing: You need to sample a large interval in order to have ANY confidence on employee productivity data. So, it is better to allow employees to just be adults for longer periods of time (say, a sprint, or even a month) and measure them by their macro output and only act when they need you (aka be a real manager).

EDIT: Oh, and one more thing, if you don't have good metrics in place (like every programming shop ever), letting the OWN WORKER pick their metrics and give their own deadlines. This is a completely valid approach and will weed out under-performers like crazy.

EDIT 2: If you don't actually act on those macro metrics and pretend everything is fine, your team will resent you start looking for other jobs.

One of my manager once told me: Sometimes you are unproductive all week and then you push out two weeks of work in a day.

It's not always this extreme, but I would be crushed, when measured on a daily basis including some very unproductive whole weeks.

+1

Especially in a discipline that requires solving new problems.

One can appear unproductive, because it is hard to get a metric on thinking about a problem.

It might take a few days of writing junk then deleting it, no commits coming for review, whiteboarding the problem, slowly learning best how to solve it. Then a week later light bulb comes on and you can solve it.

Code then gets written at high speed. And your manager thinks you're lazy, but worth keeping you around because when you work it's fantastic.

But really, you've been working all along. That writing of code was only possible because of the time it was took to think about a problem.

Uggh, reminds me of daily “standups” where I have to make yesterday’s work sound important even though half of the days are a firefighting crapshoot/standstill.

If it's a firefighting crapshoot/standstill, then that's what you say at the standup. That's exactly what it's for; not to justify how you spent your time, but to share problems and delays.

Yes, but it has to be presented in a manner to avoid blame, which is exhausting over the long haul.

I agree with you, just want to call out a pet peeve of mine: in the boiling frog experiement, the boiled frog had its brain removed [1]

https://en.wikipedia.org/wiki/Boiling_frog

I've seen the boiling frog analogy used on humans in regards to climate change... Knowing the frog had it's brain removed somehow makes it seem even more analogous than less really.

Well, that certainly colors the results.

I feel the best use of metrics isn't trying to get your teams average time with a ticket opened down 10 minutes. That rewards people who are the best at meeting the metrics which aren't even nessecarily the best employees. They're generally ones that can do easyish jobs quickly and efficiently.

Metrics have rapidly diminishing returns yet companies seem to get addicted to them in ways that make them sick. I think of stack ranking. You do want to use SOME metrics though otherwise popular low performers will get promoted and unpopular high performers will get jaded.

It's also very appearant how the lower on the payscale jobs are more metric driven than the higher on the payscale ones. It comes off as a low-cost low-effort system that managers would never willingly impose on theirselves.

Reply on news.ycombinator.com