16 comments

[ 4.5 ms ] story [ 50.4 ms ] thread
At Linux kernel, they just write code, test and send patches.)
Do they get paid to do that? If no one got paid, there would be almost no need for performance reviews.
Because if you don't get paid in $$$, you're free to do shoddy work...?

Crap work has costs that far exceed the amount spent employing the person who produced it.

Being honest (with yourself and others) in what you do is enough.
At the end of every week, month, and quarter, individuals measure themselves against their OKRs to evaluate performance.

Lee Iacoca used the same technique when rising through Ford's managerial ranks. He met with his direct reports quarterly, reviewed their progress over the past 3 months, and set objectives for their next 3 months. Then both schedule the next review and sign all the paperwork. He required that his direct reports have the same quarterly reviews their subordinates, recursing down the chain.

This, Iacoca's autobiography says, not only makes sure both sides know what was agreed to, but gets them talking to one another, and at least communicating on progress. Managers need to be part of the team too -- their role is to remove roadblocks preventing the the team from shipping.

The danger of OKR's/PKI's and other "measurements" is that it's easy to do management-by-numbers. Management-by-numbers is to great management as paint-by-numbers is to great painting.
> Peter Thiel said PayPal once rejected a top­-notch engineering candidate because he said during an interview that he liked to play “hoops,” and a PayPal engineer does not play basketball, much less “hoops.”

> Carwoo is a company that’s a little weird, so they ask every interviewee how weird she thinks she is on a scale of 1 to 5. There is a right answer. 3­-4 is the sweet spot ­­a weird person who is self-­aware.

I'm sorry, but these are just bizarre. Company culture is great, but this just sounds cultish and even cargo-cultish.

It strikes me almost as strange that we have anti-discrimination laws for hiring practices concerning races, sexual orientations, religions, etc... but companies are free to discriminate against people who play "hoops" (or don't), or are either insufficiently self-identified "weird", or overly self-identified "weird".

Well you just can't legislate against human stupidity...

Personally I think these laws are likely fundamentally flawed anyway, as hiring is subjective.
Just because hiring is subjective doesn't mean everyone is equally subjective with respect to race/gender/etc.

I mean, aesthetics are subjective too but most people prefer not to have poo on a pedestal in their living room.

You can find strong and widespread biases (e.g., against fecal art) even in a subjective domain.

Not what I mean. I am thinking in terms of proving discrimination etc.
We have those laws because there are major societal issues we are trying to address, and because we have decided those factors are not material to hiring.

You shouldn't legislate against human stupidity. Meanness, though, is a fine target for legislation.

The same article he linked to says that PayPal had trouble hiring women, so... you know, I'm not sure legality came into the equation.
I thought those were rather lame. Even most of my _friends_ don't have the same interests as I do.

Plus, in my basketball playing years, that was a great example of teamwork.

In my experience, a happy team, is a good team. We tried a lot of different interviewing tactics and had hit or miss results. Now our interview is just going out for beer or coffee with the candidate and bullshitting for a few hours.

If they aren't smart or not fun to hang around with it usually comes out in those few hours. We've had 100% success rate.

(comment deleted)
To me, this is a blog post chock full of anecdotal evidence to support a determined, central, dare I say biased, thesis. I find it continually amazing that people don't see this, from simple blog posts like this one, to now-famous books like "Good To Great" and other airport middle-management best-sellers.

Not that anything is wrong with that. Obviously, if you are interested in other people's opinion in this domain, it's likely you feel you are doing something wrong. And, trying different methods can jog whatever dynamic you are in that you feel doesn't work.

But, it's a big step from that to some of the opinions-as-fact (and eventually into "methodologies", "movements" and "manifestos") that sprout from such starting points.

I don't think anyone has the one magic bullet for "best. management style. evar."