Hacker News Reader

Comment by Daviey | Hacker News Reader

When looking at an employee, measure their output.. not their input. I'm honestly not concerned about the amount of hours put in... it could be 4 hours hard work.. but if they deliver on targets, then i'm happy.

I also want to know that they are obtainable during work hours, attentive in meetings and are available for brief interaction out-of-hours if needed.

I'm less keen to pay people to keep the seats warm, and moving the mouse just enough to stop the screensaver kicking-in.

I'm a non-smoker, but I do not mind people having smoking breaks... but some people get really upset by this. If we are doing physical or menial jobs, then productivity is reduced.. but for intellectual jobs, we should simply measure output.

(Oh, and i guess /YOU/ are probably being paid by someone to read this comment.. instead of working)

Worth reading: http://mentalfloss.com/article/74710/how-much-time-do-we-act...

View on HN

> it could be 4 hours hard work..

"Targets" are a pretty low end way to think of such an open ended job.

You can think of your job as "They asked me to do X and I'm going to do X and after that the rest of my 40 hours don't matter." Or, "I'm going to do as much as I possibly can in 40 hours." From the start of my career I always thought of it as the latter. I don't understand the mindset of the former in a professional job. Maximize your value to your employer or you'll be the last one to get promoted and the first one to get laid off.

There's always a backlog of defined tasks and there's always open-ended opportunity to make progress in ways you can think up yourself. Do you want to sit around waiting to be told what to do or engineer your own path?

You can think of your job as "They asked me to do X and I'm going to do X and after that the rest of my 40 hours don't matter." Or, "I'm going to do as much as I possibly can in 40 hours."

The healthy answer lies somewhere in the middle.

Burning out trying to overdeliver and prove your worth is no better than underdelivering and slacking off.

There is a balance. Exceeding your employer's expectations is of course a good thing as a salaried employee. That's how you build a highly successful career.

But one of my expectations as a manager is that you manage yourself so you're still around 3-5 years from now and don't need to take a break for mental health reasons.

> But one of my expectations as a manager is that you manage yourself so you're still around 3-5 years from now and don't need to take a break for mental health reasons.

You're a good manager.

> Do you want to sit around waiting to be told what to do or engineer your own path?

It can be very difficult if not impossible to engineer your own path in most environments these days. These days most devs are simply trading time for money. If the backlog board is empty, they're on retainer waiting for it to fill up.

Because of the eternal effort of the enterprise to commoditize developers they're given no decision making power. So even though they continually identify work that could be done to improve product or process, they know it's a crap shoot to follow through on such work as it's more than likely to be rejected for one subjective reason or another.

That's not been my experience over a 20-year career. On a given day I might not have that freedom, and I've had ideas get shot down. But over the long run it's always been true that if I have extra time I can use it to my advantage one way or another. This feels like a pretty universal rule about work. If I worked at Chipotle I'm sure I could still find a way to distinguish myself over my co-workers.

If you take the advantage you gained from being recognized, vs the time spent towards that effort (including things shot down), is it worth it?

The recognition must have a high return to make up for all the time wasted on rejected initiatives.

Are you asking if tens of thousands of dollars in salary, and increased equity, and promotion to roles of greater responsibility, and increased job security, and recognition, and the freedom to design my own role are worth a lot to me? Yes, they are.

You claim to have 20 years of experience, yet this reeks of "Just world" naivete.

Are you guaranteed those things if you work better than the rest? No, external factors can nullify all your dedicated work.

What would you do if all you had to show for your effort was that employee of the month picture on the wall?

First of all, "greater responsibility" is not an advantage, it is a liability you accept in return for other advantages.

If the delta is worth "tens of thousands of dollars in salary", consider what your free time spent on other things is worth. It might not be.

But "freedom to design my own role"? Where do you get this?

> First of all, "greater responsibility" is not an advantage, it is a liability you accept in return for other advantages.

This is an interesting point. We tend to conflate "greater control" with "greater responsibility", but we shouldn't. The first is a benefit, the second is a risk. They often go together, but having had a job with lots of responsibility and no control, they certainly don't have to.

The personal fulfilment from contributing to a constantly-improving work environment is what makes it worth it.

This seem to ignore something in my original comment: What about the wasted effort?

What wasted effort? Sure, not all of my initiatives gain traction, but to expect otherwise would be unrealistic. Perhaps I've just been lucky to work at companies that are receptive to personal initiative, improvements and ideas.

Regardless, I would much prefer to have a working environment where I'm engaged and involved and making an impact than where I'm only grudgingly doing just enough to get by, even if it doesn't lead to recognition and rewards. Although, in my experience, it does.

In a healthy organization the engineer can suggest improvements to Product and help them prioritize the improvements.

Agreed, and I've worked in such environments. However, its my thesis most "dark matter" engineers in "the enterprise" are in an environment I originally described. And that's not accounting for Government organizations which I imagine are even worse.

> I'm going to do as much as I possibly can in 40 hours.

However the inverse isn't true; your employer doesn't pay you as much as they possibly can for 40 hours work. They pay just enough.

So why not work just enough? It's meant to be a balanced exchange.

That's a narrow view and inaccurate over the long run. I make more money than my colleagues who only do what they're asked to do. And if my company has layoffs I'm one of the last who would be let go.

In many layoff scenarios, the first to be let go are often the most expensive ones (the ones who make more than their colleagues who only do what they're asked to).

I've seen the best people let go only in cases where an entire department was laid off as a whole unit. In cases where people were selectively let go I've never seen it be the top performers.

>I make more money than my colleagues who only do what they're asked to do.

That's not the case everywhere though, which is the observation that TFA and the parent were starting from. Obviously if you do make more it's ok.

>And if my company has layoffs I'm one of the last who would be let go.

You'd be surprised how many people thought that and got the short end of the stick.

">And if my company has layoffs I'm one of the last who would be let go. You'd be surprised how many people thought that and got the short end of the stick."

It's a delusion that helps people sleep at night...

Since it's both logical and fits with 20 years of observed experience it's reasonable for me to act as if it's accurate. Despite exceptions and anecdotes to the contrary.

The times I've been at a company where shitty politics supersede reasonable practices, I've left. And having a resume with greater achievements makes this easier than a resume with fewer.

>Since it's both logical and fits with 20 years of observed experience

The famous example Taleb uses in his book is the Thanksgiving turkey.

"Consider a turkey that is fed every day," Taleb writes. "Every single feeding will firm up the bird's belief that it is the general rule of life to be fed every day by friendly members of the human race 'looking out for its best interests,' as a politician would say.

"On the afternoon of the Wednesday before Thanksgiving, something unexpected will happen to the turkey. It will incur a revision of belief."

http://www.businessinsider.com/nassim-talebs-black-swan-than...

[Posting from a throwaway]

You're exactly right, in an unexpected way though.

I was working at a somewhat successful BigCo. (makes HN frontpage every other month or so) a few years ago. Then came the time to discuss raises that year. I did more than was required of me, helped other teams, rewrote a bunch of legacy code in languages that were not part of my job description, etc.

What did I get? A sub-inflation raise in a year where the company's revenues increased by 70%. My closest coworkers received about the same. All because we worked on infrastructure, deemed a cost center.

So next year I coasted and put in the bare minimum while working on side project and taking freelance work. Got average to below average marks when the special time came. Then got asked to work unpaid overtime to "fix my mistakes". I quit on the spot. It was a huge relief.

Now I work for the man in the mirror and he's the best boss I could have ever wished for.

Only when working hourly with a guaranteed contract does "engineering your own path" necessarily result in extra compensation. A rational person will put in effort up to (but not past) the point where that marginal extra effort results in an a commensurate expected marginal reward.

Assuming you're rational, say you're paid $P for Q output. If you thought you had a 75% chance to get $P+10% for producing Q+5% output, you'd be right to work harder and produce that extra output. If, on the other hand, you thought you had a 10% chance to get $P+100% for producing Q+50% output, you would not accept that.

Most employees aren't capable of 40hrs/week of intellectual output.

There might be some technical accuracy to that but it's a lame and self-defeating thing to rest on as the basis for your career strategy.

I'm sitting here at work typing this reply to you. It's not about using every hour, it's about fundamentally seeing your output as malleable instead of fixed based on your manager's "targets" and seeing yourself as responsible for your level production.

Most of my time (and much of my team's time) is spent on projects that were my idea. It wasn't that way at the start of my career but I've always works towards taking control.

what you just said is completely orthogonal to the comment you're replying to.

self-directed vs. manager-directed is a separate concern from whether or not a programmer can be expected to put in 40 hours of high focus, high concentration technical work per week.

in my experience getting 20-30 high focus hours in a week is a very good week. the rest of the time has to be used for things that are much less focused. like arguing about this shit on hacker news, for example. it's not even an either/or scenario. you need to do both. just like when breathing you have to both inhale and exhale.

I don't know you but thank you for treating your employees as professionals by setting expectations and letting them deliver however they see fit. I've worked at places where they "treat everyone like a hammer" so to speak and if you are delivering but you're doing it like a screwdriver you're "not a team player". Simple way to describe it is a bizarre emphasis on culture fit over productivity.

So long as, in being a screwdriver, you don't piss off and alienate all the other hammers.

Speaking for myself, I'm done with primadonna developers who think they can shut the door and code alone while running roughshod over their colleagues, then after a couple of years take off and leave behind their mess for others to sort out.

Development is a team sport. If you don't buy into that you won't be working at my company.

True but what if there's skill gap between team members.

The skilled devs should be coaching the less skilled devs, including pairing on work, code reviews, etc, in order to spread the skills around.

Meanwhile, management should provide opportunities for training and so forth to develop their talent.

Not only that - management should actively provide incentives for the more skilled devs to coach the less skilled ones, otherwise it just becomes extra work.

Then the higher skilled members should devote a significant amount of time getting others to level up. It is a team. Management should have structure in place to incentivice it. If they don't have a structure in place then the high skilled members should devote time to make a proposal and pitch it. Find a way to win as a team.

haha everyone says "just train your employees more!". I do like the the "get them to level up" mentality.

BUT... it won't work. We're all on a spectrum of ability. Maybe more so, we're all on a spectrum of ability spectrums. Training, education, these aren't the answer to "my devs aren't as good as that one". You take the ability you have, your team has, and you manage and apply it the best way you can. Most often, that doesn't mean having your top developer spend her time training everyone else.

It means you give that top developer the protection and time she needs to get her shit done. If she happens to be the type that gets energy from training and helping others, THEN you have her train others.

In computer engineering terms, you have to balance your control path as well as your data path.

It means you give that top developer the protection and time she needs to get her shit done.

But not to the exclusion of all else.

Again: a team is a team. I, as a manager, need the team to functional optimally. If one developer is a 10x hotshot, but she makes the lives of everyone else substantially worse, I don't give a crap how much code she can write, she can find another team.

The top developer doesn't get to be a silo, a dictator, or a troublemaker. As a high-skilled individual, they are expected to produce code, collaborate with team members effectively, provide mentoring and training, and generally lead by example.

Again, if they can't handle that, they can find a company where their style is a better fit.

> But not to the exclusion of all else.

No, that's where you're wrong. If that developer is really your top producer, the bottleneck of production, you isolate and protect her exactly to the exclusion of all else. She becomes the worker that cannot, will not, be bothered. It seems counter-intuitive, but read on industrial engineering practices around optimizing an assembly line and you'll see it shown all around.

To get your team to function optimally, you must protect the core assets. And given that those assets are people, they have to feel protected. In the end, the rest of your team exists to support whatever it is that actually creates value. If that happens to be one developer who has their hands in 50% of the code (that others couldn't support if they wanted), then the rest of your team exists to support that one developer.

Ideally, though, no team is so one sided. There are no true 10x developers. And those that exist utterly fail at the rest of team management. At the end of the day, give your workers work they can actually do.

It seems counter-intuitive, but read on industrial engineering practices around optimizing an assembly line and you'll see it shown all around.

Development isn't an assembly line, and the fact you'd use that analogy says a lot...

If that happens to be one developer who has their hands in 50% of the code (that others couldn't support if they wanted), then the rest of your team exists to support that one developer.

Oh heck no.

If one person or only a few people are creating most of the value on the team, the team is dysfunctional and you've exposed yourself to enormous risk. If those people leave, fall ill, get injured, or go on vacation, you and the team are screwed. If you, as a manager, have put yourself in that situation, you have failed.

All work is an assembly line. It's not a value statement, and it doesn't take the art away from the work. Work is work is work. Developers just have their head shoved too fair to get away from feeling like special snowflakes.

You're talking about risk management, not individual worker ability. Yes, it is a poor idea to have any one portion of a system maintained by only one person. So don't. It is a poor idea to have only one person who understands the system architecture. So don't. It's a poor idea to not recognize the different strengths and weaknesses of all your workers regardless of context. So don't.

I wouldn't want to work for you.

Good to know.

this is the kind of batshit insane "logic" that produces cultures like the one at Uber.

The idea of protecting your most productive asset so that it can continue to be your most productive asset instead of burning out solving problems its poorly suited for is "batshit insane logic" that prevails Uber? No, no its not.

Just because I label development an assembly line doesn't mean treat your developers like machines. They are machines, human machines. Meaning they have feelings and a whole host of things that must be managed in addition to keeping them well oiled.

"Just because I label development an assembly line doesn't mean treat your developers like machines."

"They are machines, human machines."

Contradiction.

no, people aren't machines of any kind. software development is more like creative design than assembly line work. your metaphor is flawed and the way you talk about the process and the people is misguided.

The employer should endeavour to close those gaps via training or even just advocating things like more pairing. It shouldn't be acceptable for a developer to silo themselves off from the rest of their team (except maybe a non-artificially-short-deadline or start-up type environment).

How do you measure what they should be getting done objectively? If you think a job should be done in a week and it's not, it seems they should then be fired because they're not "getting the job done".

We have that discussion a lot. How do you communicate that in a constructive way? How do you provide ongoing feedback around it?

I think it odd that some companies that say they practice Agile/Scrum still track hours. If the each team member commits to a certain amount of work, it doesn't matter if they put in 60 hours or 16 hours doing it.

Of course, there's meetings and fires, and so the dev should be available for that, but, in all, who cares about the number of hours?

If the each team member commits to a certain amount of work,

If you're committed to Agile, I don't understand why you're talking about individual commitments at all.

The team commits to sprint scope. The team succeeds or fails in meeting that commitment as a group.

Beyond that the team should be free to figure out how the commitment is met.

We have one team that works 10:30 to 6:30. Another that works more traditional 9-5 hours. One is largely remote. Another is co-located. They all produce high quality work at a solid, and most importantly consistent, rate.

If we tried to force some work pattern on them, I doubt they'd be as successful.

Hours let you develop much better estimation models. Agile/Scrum isn't a silver bullet and it is valuable information for any development shop.

https://www.joelonsoftware.com/2007/10/26/evidence-based-sch...

> certain amount of work

This is the crux. How does the business decide what's a reasonable amount of work to get done in a sprint? If the dev in the OP has time to work on his own stuff, why shouldn't the business ask for a bigger commitment in the next sprint?

They can; that's why they track a team's velocity.

How do you measure output and what do you compare it to?

Measuring output sounds like an ideal way to evaluate employees, but also sounds pretty challenging in and of itself.

To be clear, identifying a poor engineer seems simple enough. But how do you tell the difference between a bright engineer working on a hard problem and a bright engineer being lazy?

How long a task "should" take is often totally unknown. After all, if a programming task is so routine and well defined that it can be accurately estimated, that is probably good indication that this task can be automated, so we shouldn't be regularly doing such a task.

Yeah, most employees read reddit or HN in between doing work tasks in order to reset their minds a bit.

If the employee is doing external coding instead of surfing reddit on their 'mental floss' time, then that's probably a good thing and its making them a better programmer.

There's issues of course if they're making competing products, or using company proprietary IP, or making money on the side.

But this may come down to the fact that the biggest issue is that you've got an open office plan, where all your employees are snooping on each other. And if this employee was doing this side-coding behind closed doors (or at home) that nobody would care as long as they were hitting their targets.

And with that I'm off to grab some coffee and get back to work...

>I'm a non-smoker, but I do not mind people having smoking breaks... but some people get really upset by this.

I was in a group where the non-smokers "saved" their smoke breaks and we all went to breakfast on Friday mornings. We'd take about a 45-60 minute break and the smokers were not allowed to come.

well if OP didn't have issues with his performance, then they wouldn't have asked the question. OP's running a dev team in a small startup, so my expectation would be "no free time during work until we stabilize"

Yada yada.

Heart that so many times and it is crap.

Only because i work better than some avg and my output satisfies you, you pay me for working 8 hours and not for something else. I'm going home after 8 hours regardless if i'm done with a task i thought i will finish, why should i leave/slack after 4 hours for a task someone thought it would take 8 hours?

you pay me for working 8 hours and not for something else.

Actually, if you're a salaried employee and not being paid hourly, that's objectively false.

A salaried employee is paid to ... Well, it's up to the employment contract to decide.

Personally, if I hire someone on salary I expect them to estimate level of effort honestly to the best of their ability and complete work at a rate I consider acceptable. If I lowball that expectation that's really my problem.

And the reality is most employees would be dissatisfied with that situation, because most people want to do meaningful and challenging work. So if someone completes their assignments early and ends up filling their time with side projects, it's because I've failed to give them work that challenges them.

The only thing I expect from the employee is that they tell me when they're done so I know what's going on and can decide what I want to do about it (and that should mean working with them to give them more satisfying work so they don't leave out of boredom).

You are far above average if you can be in "the zone" for 8h a day. The rest of us will just have to "slack" while you do the hard work.

> I'm going home after 8 hours

Because you don't have competent technical management.

My team comes and goes as they see fit, we have the highest profitability margins of any of our departments so noone in HR bothers to tell me anything.

Edit: However, with that comes 3AM phone calls where I expect them to act like adults and do what needs to be done, so I'm not perfect by any means.

Reply on news.ycombinator.com