Hacker News Reader

Comment by akozak | Hacker News Reader

Number of hours still seems like the wrong unit of measurement to me. Shouldn't companies care more about productivity?

If I am still productive at 60 hours and I am being paid a salary, then why not work me 60 hours? Employers care about getting things done, employees care about how their time is spent. Employers need (~)happy employees. I think hours is the right measurement here.

Perhaps the issue is exempt employees? I don't see why I shouldn't be paid hourly.

Yep. My "salary" is really just my hourly rate times the number of hours in the year if I work 40-hour weeks. I can get 50% more pay if I work 60-hour weeks.

It's easy to measure workers by how many hours their ass is in their seat per day, so that's what gets measured and that's how we're judged.

I'm tempted to say I'd rather be paid per bug/feature, with the pay rate increasing based on the complexity of the bug/feature, so that when I think I have enough money I can say no to additional work without repercussions -- but workers in other industries organized against [piece work](https://en.wikipedia.org/wiki/Piece_work) for a reason.

Per bug you say? I'm gonna write me a new minivan this afternoon!

http://blog.klocwork.com/wp-content/uploads/2009/10/dilbert-...

A rational company would, but in practice lots of companies either don't know how to measure productivity, or they are confused enough to believe that long hours will result in more output.

So as an employee you want to ensure you have a hard limit on how many hours you're expected to work (40 for regular jobs, or as in this case, less than that).

And what's a good productivity metric?

https://www.folklore.org/StoryView.py?project=Macintosh&stor...

That'd be very job-specific.

It's probably misguided to aim for highly quantitative measurements. If productivity is measured as output per input, you have a lot of flexibility in how to define the relevant output - e.g. hitting well-defined product goals or milestones (versus e.g. lines of code, which seems wrong).

Widgets created

commits per day

I agree - google/check out something called the "Results Only Work Environment" or "ROWE" it's a great framework for this.

They care about productivity, but that's more difficult to measure directly, so they need some proxy metric for it. Hours worked seems like a reasonable metric that can be applied to the broader workforce.

Right I get why - but it's crazy to think that the vast majority of companies give up measuring the thing that everyone knows matters, especially when gearing your incentives around it should be better for everyone. (except low productivity workers I guess?)

Can you even define productivity? On the surface it appears as easy as getting work done, which if only a single person is on the team that may be enough. But once you add a group of people together, mapping a teams productivity to a single metric that can be applied to each individual becomes much harder if not impossible.

Basketball has had this problem for a long time. The plus/minus metric was invented as an attempt to address all the things players do while on the floor, but do not directly relate to an easy to capture metric. It also tries to address players who put their own personal productivity over that of the team.

Reply on news.ycombinator.com