[pmarreck]

This is why, if I ever manage, my focus will be on timely deliverables and not on time-at-desk. Give me a day when you think you can deliver something, update me as soon as you realize you can't ("manage expectations"), and do whatever you want with the rest of your time...

[munchbunny]

I think this is only half of the picture. Here's a formal framework that talks about it.

https://en.wikipedia.org/wiki/Task-oriented_and_relationship...

As an example, when you are deciding whether to promote someone, you can't do it on tasks completed alone, because some of that decision has to be about (1) how well the person works with teammates, and (2) whether the person's peers believe this person is competent enough for their promotion. If you don't get those two considerations right, you compromise people's trust in your competence and the fairness of the promotion system. They're fundamentally relationship and skill development issues and not deliverables.

If you want to be a good manager, you have be fluent in both approaches, and you have to learn when to use a task oriented approach vs. a relationship oriented approach.

[mleonhard]

Thanks for sharing this.

[majormajor]

That sounds lovely in theory but gets tricky quick with employees who give you estimates that seem too high, and then don't even meet those.

And then you have to figure out if they need help, if they're over their head, if they're being lazy, if they're getting interrupted by stuff that's not on your radar, or WTF is really going on, and if (a) they're not telling you and (b) you have no visibility into their day (say they're remote), then it gets really hard fast.

And the more insecure they feel, the less likely they are to directly tell you they need help.

So as manager: don't joke about firing, don't joke about time at desk, don't joke about performance.

But as employee: if you get your shit down, and have talked to your manager about it, and you're both good with the amount of stuff getting done, you shouldn't be concerned with taking a day off here or there, or leaving early, or working from home. But be aware that a lot of performance monitoring stuff isn't for you, and not everyone is doing as well as you are.

[danenania]

"estimates that seem too high, and then don't even meet those"

This is pretty much the definition of an estimate in software.

[amznthrowaway5]

Since the estimates are almost always off, it's a perfect way for the managers to reward the people they like for political reasons, and punish the ones they don't.

[amznthrowaway5]

> That sounds lovely in theory but gets tricky quick with employees who give you estimates that seem too high, and then don't even meet those.

How many managers are competent enough to judge that kind of stuff, and whether or not any excuses produced are reasonable? Even in the technical manager cases I've seen, maybe 1%. The people who constantly bs excuses are often the ones rewarded, as opposed to the people who actually focus on the work and take a little longer to get it done properly.

[pc86]

> Even in the technical manager cases I've seen

You say this as if technical managers are better managers? In my experience the opposite is almost always true.

[amznthrowaway5]

At least for engineering manager positions, there shouldn't be any non-technical people. How are non-technical managers supposed to allocate work and judge how well their technical employees are performing.

[moftz]

That works until some consultant sees that one employee seems to be pretty quick getting their tasks done so management is told to have even more work assigned to them. Why give someone 30 hours of work if they are being paid to do 40 hours? The demand of work is going to inevitably fill the supply of time. As an employee, the proper way to deal with this is just find the proper amount of work you feel comfortable doing in 40 hours and make that your permanent pace. Work too fast and you'll be expected to perform like that forever. Work too slow and you'll be closer to the chopping block whenever the time comes. I'm not suggesting to be lazy and goof off as much as possible to fill the 40 hours but I am saying to expect typical management to be more concerned with how much more you could be doing rather than how much you already do.

[whiddershins]

One way that fails is when you are trying to nurture talent that really isn’t where it needs to be yet.

[pmarreck]

Maybe measure estimate accuracy over time, then, and reward improvement there.

Surely, as one gets more skilled, their estimates will tend to zero in on accuracy more

[whiddershins]

I’m just saying when someone is learning, I really am looking for them to put the time in for a while.

Skills tend to increase in nonlinear jumps, so I’m not looking for a targeted increase. I’m looking for someone to discover what they are actually capable of which is often more than they anticipate.

A great way to get there is to immerse yourself.

[umpc]

That is more or less what my current manager does. He's the best that I've worked with.