10 ms·
In the mid aughts, I worked at a company that put ROWE (results oriented work environment) into practice. This approach basically meant you could work whenever
by burningChrome 3y ago
In the mid aughts, I worked at a company that put ROWE (results oriented work environment) into practice. This approach basically meant you could work whenever you wanted, as long as you were meeting goals and metrics. There was a list of metrics you had to meet before being eligible for it. They kept it in place for two years and nothing really changed. Devs were still meeting their metrics and the company was still doing well, but when the old VP who put ROWE in place left, the first thing the new VP did was take it away along with some other perks we had. Myself and about seven other devs left within the next 3-4 months.
I'm currently working for a large corporation who went through several cycles of trying to get people to come back to the office after C19 slowed down. After three or four versions, they finally gave up and put an optional (hybrid) model in place. It was interesting to hear the stunned VP's glowing about the increased productivity and the company had two of its most successful quarters going into and coming out of the pandemic. I'm guess seeing those results made it easier for them to allow people to work from home more easily.
Some anecdotal evidence for your retention theory. Since my current company allowed 100% WFH, our team has had barely any turn over and at least three of the teams I worked with have also have little or no turn over as well. It would stand to reason you probably have a good point about retaining people when you allow them a little more freedom to do their jobs.
- sandworm101 3y agoI'm with a large organization (government, military) that also implemented a hybrid model under covid and decided to keep it going. But not a day goes by that someone doesn't scream for it to end. We have too many people "working from home" all the time and never getting done the stuff that cannot be done from home. Is the server not responding after the recent power outage? Too bad. The guy to turn it back on only comes in Monday-Tuesday. They put the real property maintenance people on a hybrid model. No joke: The plumbers and electricians still "work from home" half the week. "Hello, this is General Smith." "Sir, ... um .. you are answering your own phone?" "Yup. My EA is working from home thursday-friday and we cannot get the secure phone to forward calls to her cellphone." "Did you talk to IT?" "It is friday. IT works remote on fridays." "How about I come to your office?" "Please do, I'm all alone here."
- vidarh 3y agoMy employer's IT is in a different country. I'm not sure which. If it matters, that's an organizational problem, not a WFH problem.
- ben_w 3y agoAt the rate things are currently changing, I won't be very surprised if non-trivial server capacity gets located in orbit thanks to some successor to Starlink. This will make remote work mandatory, even if it still can't fix all the possible issues.
- OJFord 3y agoWhy would that make remote work any more mandatory than it is with 'the cloud' or data centres?
- ben_w 3y agoGoing to space to fix them is very hard. (Only due to your comment did I realise I was ambiguous, I meant remote work specifically for the IT department would be mandatory, not everyone in general).
- OJFord 3y agoIt could still be office-based though, sure technically remote from the server, but so are the web servers software engineers develop for typically, sometimes even their development environment is. I don't think that really factors into remote (WFH) vs office working at all, at least from an employer's perspective. Maybe it makes it easier to work remotely, but certainly doesn't necessitate it: you can work from anywhere, and that can still be the office.
- HowTheStoryEnds 3y agoYour worklocation is mir and it's you that goes remote, you don't go remoting. Just like a drill rig. ;)
- nox101 3y agoI have no idea how metrics would be measured unless the work is always the same and trivial. In my job, every problem seems to take a random amount of time. It might be the thing I'm assigned is trival and done in 1-2 hrs. It might be it would be trivial but it pointed out an issue elsewhere than needs to be fixed first. It might be trivial but the CI was down for 6hrs. It might be trival but requires some other library to roll to a new version and the roll broke something else. It might look trivial but turns out to need a non trival refactor to work. I ran into that last one yesterday where I started on an issue that seemed like it should be trivial, in an object with different options than it currently has. These objects are cached by their options since they are heavy. The code that gets an existing or new one is 5 levels deep. It creates them from a factory but the info this new issue needs needs to query the factory to figure out the creation options. So, what seemed like a 10-15 min change is going to be several hours to decide how to surface or pass that factory to/from the top level the point being, from outside management is going to see different levels of work. It's not as trivial as "created N widgets an hour" nor is it as trivial as "lines per hour" I feel like there could be lots of unintended consequences with people only taking the smallest tasks or whatever to avoid taking tasks that take longer or to find ways to account things so it looks like they're doing more, etc
- ozim 3y agoThis is exactly why I oppose any dashboards with amount of completed tasks per employee. This just gives wrong ideas to the team.
- PurpleRamen 3y ago> I have no idea how metrics would be measured unless the work is always the same and trivial. Don't count the result, but the actual work itself. Meaning, you need to document every little step you do, every command, script, search you have. Write down your thoughts and conclusions, and so on. But this is very annoying, and barely anyone is well-equipped for doing this. And most people have some natural defensive for exposing themselves to this level. Though, it would also be helpful for yourself to be able to review your work. So maybe some balanced middle would be a way to go? Maybe if tools would be designed in a non-annoying way to support this, more people would be willing doing something like this.
- hn_throwaway_99 3y agoI'm really curious as to how ROWE works in practice for software developers, because the I'm extremely dubious. In software development, there is usually an "infinite amount of work" to be done, so the important thing is to prioritize well as a lot of that work will never get done. But what does it mean to say "you've met all your objectives"? If you say "great, I finished all my objectives in 20 hours this week, see you at the beach", this company will eventually get killed by competitors who are just as productive yet work 40+ hours a week. But fundamentally, the idea of "you've met your objectives, now you can go home" just feels like a fantasy in the world of software development.
- p1esk 3y agoWhat do you do during sprint planning meetings if you don’t know your objectives?
- hn_throwaway_99 3y agoBut it's not like you complete your objectives and then you say "great, we're all done, let's go home." Every company I've ever worked for had a backlog a mile long, and we knew we'd probably only ever get to 20% of it or so. The whole purpose of sprint planning is to just prioritize the most important stuff to work on for the next sprint, but if the teams capacity increases you'd pull more stuff in.
- p1esk 3y agoNo, the purpose of sprint planning is to specify (more ot less) exactly what you plan to do this sprint. Usually the best case scenario. You make your best effort to finish the things you outlined, sometimes you manage to do it, sometimes you don’t, but why would you ever try to do more than you planned to do? If you’ve done the tasks, you’re done, period, go relax, grab a beer, wait till the next sprint. You’ve literally done your job. Why in the world would you want to “pull more stuff in”? That’s how I’ve done it, and have seen it done in the last 7 companies I worked at. Your reasoning is so foreign to me. The only way it makes sense if you are speaking as a manager.
- gnicholas 3y agoDo you think the lack of turnover might be affected by the current macroeconomic factors? It feels like lots of people have been holding onto their jobs for the last 18 months for this reason.
- gymbeaux 3y agoIt’s interesting and encouraging to see this working in the real world. I think most “leaders” are concerned about having to come up with what people, especially software engineers, should be getting done in X amount of time, which I sympathize with… but clearly it’s doable. Perhaps the concern is that high-performers will produce less, and low-performers will be canned, only for the high-performers to have to pick up the slack, or there just be a net negative in total work produced in X amount of time because some engineers have been fired for performance but there’s no one to replace them and others aren’t allowed/able to work more to compensate.
- nitwit005 3y agoThat tale of a policy being scrapped the moment new leadership shows up is typical though. Leadership churn is often high these days, so workplace experiments or innovations tend to just vanish, unless they become popular across multiple companies very quickly and become the new norm.