(1) all employees are trusted and unmeasured, but you have to tap people on the shoulder every once in a while to confirm that they're on track. Naturally, this is easier if everyone is on-site.
(2) everyone is accountable for producing of tickets, and you can check that everyone is at least doing some work. Check-ins can happen via comments in tickets.
(3) Everyone just does whatever they think they're supposed to be doing, and the manager only finds out something is wrong when the employee volunteers the information or the project isn't delivered on-time.
(1) is the common case, (2) requires time and skills many managers don't have (and creates @#$% Jira commentary form the peanut gallery), (3) requires building an excellent team with years of mutual trust between them, and still goes wrong all the time.
Seeing you at your desk working, and asking a casual question or two at the water cooler is by far the easiest (laziest?) way to make sure things don't go too far sideways.
I don't get it - shouldn't one-on-ones and regular progress check-ins (be those standups, metrics (This is your #2), whatever) give you that information? None of those are easier when on-site. In fact, given today's move towards open offices, any form of 1:1 collaboration is easier remotely where you don't have to fight for precious meeting room space.
I think the core problem is that you're trying to prove "everyone is working" when we should be interested in "the work is getting done". Seeing that everyone is working doesn't actually mean progress is getting made. If you have to take steps to answer that question anyway, the first question is fairly pointless.
Also had managers that were basically invisible but kept a great crap umbrella for us. Loved them too. Ended up working a ton and getting enormous amounts of work done because trust and autonomy are huge.
I think the worst situations are when managers are too insecure to be honest with direct reports. That’s decidedly uncool and super ineffective.
Also, it seems managers get so overloaded that 1 on 1s and stand ups end up being mostly a waste of time because it’s too hard to remember all the little details.
So remote or in office isn’t really the issue. In my view, remote is superior due to productivity, flexibility and happiness, but it does require everyone to do regular video chats and pair programming.
Ironically, the in office folks have the hardest time with this because they feel they need to get up from their desk and find an overly booked conference room as you mentioned.
Simple solutions are context dependent, but could be along the lines of a) work remote if you have the discipline, b) asynchronous slack based standups, c) report status as you go in your tickets, d) every day or two peers and managers sync up 1 on 1 or in person depending on their preference, e) have clear goals as a company and a team, f) have good technical product people writing 1-3 day user story tickets including completed designs if it’s ui work, g) have good tech leads writing 1-3 day technical improvement task tickets, h) vote as a team on contentious technical decisions, then disagree and commit, i) generally chill out a bit and let work be enjoyable.
At least that’s a start...
> Also, it seems managers get so overloaded that 1 on 1s and stand ups end up being mostly a waste of time because it’s too hard to remember all the little details.
I'm not sure how the first quote doesn't count as a 1:1 - the goal isn't to have a highly formalized process, it's to have communication. Some about the specific work, some about general work and workplace, but communication.
Total agreement about the overloading of management though - I haven't had a boss in years that didn't regularly have most of their schedule booked and often double-booked. I had to fill in for my manager while he was on vacation for week and it reaffirmed my lack of desire to do management - and I only had to deal with pieces that couldn't wait a week! I sometimes think the recent shift to focus on regular management 1:1s is all about reclaiming enough time to actually know what their team is doing.
Yes, you’re right, I guess those were a form of 1 on 1s :-)
> Total agreement about the overloading of management though - I haven't had a boss in years that didn't regularly have most of their schedule booked and often double-booked. I had to fill in for my manager while he was on vacation for week and it reaffirmed my lack of desire to do management - and I only had to deal with pieces that couldn't wait a week! I sometimes think the recent shift to focus on regular management 1:1s is all about reclaiming enough time to actually know what their team is doing.
Absolutely! It’s wonderful to have empathy for this! :-D
> I don't get it - shouldn't one-on-ones and regular progress check-ins (be those standups, metrics (This is your #2), whatever) give you that information? None of those are easier when on-site. In fact, given today's move towards open offices, any form of 1:1 collaboration is easier remotely where you don't have to fight for precious meeting room space.
This is not quite the case everywhere. We have agile 5-8 person offices and people are still supposed to walk out for longer phone calls occupying one of the smaller meeting rooms.
But I actually like that different teams can mix a lot easier.
I've heard this claim for open offices a lot. I never understand it. It's not like I'm AGAINST easy collaboration. But I spend (or try to spend) more time on my own, and everyone "collaborating" around me and expecting that I can turn off my peripheral vision and hearing with a brain that is hardwired to NOT ignore those things just makes the majority of my time worse.
Plus, in the last 20 years and 5 companies (spanning 8 offices if you count office moves) I've never had adequate meeting room space (defined as: We need a space to spontaneously talk without disturbing others, can we find it trivially?) for more than a 3 month span after moving to a larger building. Hardly scientifically conclusive, but personally persuasive.
So I'm all for collaboration between people, including between teams, but I don't understand sacrificing the REST of the time in the name of that one thing.
My team shares an open office with another one, working on a different project. We see them every day.
I don't think anyone knows any name of the people in the other team.
Only if you are a contractor and you've fulfilled your contract. If you are a salaried employee I expect you to keep your hopper of tasks full so you aren't sitting around idle 95% of the day waiting for me to feed you 30 minute taskers
The "goal" in a lot of cases is nebulous (i.e. "keep developing the product", "improve stability", etc). There might not be a manager feeding you tasks every time you run out. At least in my company, you are expected to figure out on your own how to continually make the product better and proactively do it, not sit around idle after rapidly finishing the last task your manager gave you.
If you are just a dumb worker that can only work when your manager fills your queue, you are not valuable to me as a salaried employee (you are more valuable as a contractor who I can call upon for a one-off difficult task that you can then complete in 30 minutes).
Here's what I believe:
If I pay you a salary, I want (roughly) 40 hours per week of productive work at your current skill level. Note: I consider sitting around thinking about things relevant to the product or company to be productive work. If you believe your level of output deserves higher compensation than you are receiving, come talk to me or find a company that will pay what you think you are worth. However, accepting a salary and then doing tasks really fast and then going idle (watching netflix, working on side projects, etc.) until your manager cattle prods you is not acceptable for a salaried employee; that would be more appropriate for a contractor where you can idle on your own dime.
- you have an employee that works faster than others, in the same pay grade.
- You are annoyed that he or she has empty queue of work
- You expect employees to look for tasks on their own when they are done
so either that employee:
- has no creative control over the work he or she is doing and cannot move forwards on their own.
- is severely underpaid and undervalued
- is under bad management that cannot fill their work queue fast enough
All of that will lead into them doing tasks at the same speed as other ones - you just turned a good, valuable and fast employee into another drone.. or worse - they'll go work for competition, and you'll have to both suffer the loss of a good trained employee and a cost of hiring someone else.
They are not the "best employees" if they are constantly idling, waiting for me to tell them what to do next. People here seem to think the ideal manager/subordinate relationship is that of a worker thread pool where the manager fills the queue and the worker threads pull tasks off the queue.
Not so. The "best employees" are ones who can examine the current state of affairs, and then come to me and propose things to work on. The ideal manager/subordinate relationship is actually one where the workers are all artificial general intelligences that learn over time what the product is, how to improve it, and proactively create tasks to work on.
If you are a rockstar but only work as a a super-fast super-efficient thread in a worker thread pool, you are less valuable to me than a non-rockstar AGI who can dynamically recognize which tasks need to be done and only occasionally needs guidance and course-correction.
If employee is getting extra works as a reward for being efficient guess what will happen.
If you're truly doing all those things in 30m, you should be running the show.
Sometimes taking the full 8h allows you to put in the packaging to confirm you've done the "hard" part.
Sure, some unknown quantity of people can just fuck around and hide from work, but again, you're not paying me because I'm a fuck off person, yer paying me because I have the decades of experience, and because you know with certainty I'll do the damn work, find more work when I need more, and will never miss a deadline on my own fault. Everyone on HN just assumes bad faith all the time, geez....
It feels good to help people, as long as it’s with healthy boundaries so as to prevent burnout.
Most tech workers get paid a bit more than that.
I did actually signed contract that actually specifies fI'll time - 40 hours.
I you wanna count weekend thinking into it, is a different question.
It's a weak manager that's constantly concerned about their reports not being productive. Nothing in the world is easier to spot than someone who doesn't produce. You don't need to optimize for that, when it happens, deal with it. Optimize toward keeping your team motivated.
i would say that it is a very quick way for employee to reach towards a new company.
I just finished working on a fully remote contract that worked that way, every day at noon there was a conference call that never took longer than 15 minutes. It worked great. A couple manager types were on the call, and they'd share any developments on the contract side of things, feedback from the customer, things like that, but mostly they just listened and took note of potential issues and made sure they got resolved (so like if you mentioned a problem one day then didn't the next, they'd ask if it got dealt with).
Don't waste a day resolve that issue now.
We have 8 people on our current project team and we can typically finish this meeting in under 5 minutes each morning. We have some people on each coast, and half the people are remote so we do it at 10:30 am Eastern.
It keeps everyone accountable to the company and each other, and because everyone knows what everyone else is working on, it makes it easier to avoid overlap and divide up tasks in the most logical way possible.
When I've seen "things go side ways" it's almost exclusively at higher levels of management when politics create goals other than "create an amazing product". Anyone who has worked both in a large company and a small startups has likely seen the difference between "my boss's goal is for us to create a best in class product" vs "my boss's goal is for him to get L_current +1 promotion".
There are no companies that have failed because the workers have finally figured out how to sneak their way out of doing work. But workers do have a hard time being productive when "productivity" is some weird game that no one will really explain the rules to.
But the logic of blaming workers who get "off track" is a huge part of the culture we need to get rid of to have successful remote work.
It's not about blame, it's about trying to ensure the team works together cohesively, and that individuals are unblocked.
I suspect there are such companies, but that the guilty individuals weren't serious professionals. Hire bad enough 'developers' and I can see it becoming a problem.
If the job is shovelling snow, I imagine it becomes one of the management's chief concerns.
If the only method for your manager to know what people are working on is to tap people on shoulders, then doing that isn’t a “bad behavior” for the manager, it’s necessary behavior.
Of cause a better solution is to have trust and plans. (Agile, scrum, waterfall, napkin sketches or whatever it doesn’t matter as long as people align during planning and have transparency into the plan). But in the absence you still need to have management in sync with what is going on.
Naturally the best pathway forwards is to look into changing your system to include all these value adding elements. But enabling remote work while management is critically dependent on physical presence to continue operation is not a good idea. And the solution is not just for managers to “give up the mindset” of knowing what’s going on below them. Trust is great when it works, and toxic when it doesn’t. The agile mantra of increasing trust is not just supposed to advocate blind trust. It’s about building a framework and working routine where we are validated in trusting each other.
I honestly think that EVERYONE should be using a ticket tracking system and just put assignments in a person's queue. I don't see why this is such a big deal or looked down upon. It's so easy to see what everyone is working on, run reports to see how long things took, have a place to keep track of notes which builds a knowledge base, the benefits are limmitless.
Not to mention the fact that if all you have to do is look at the reports to know what is going on, you don't physically need the person there which will allows them to work from anywhere.
As I see it, none of the queuing stuff is really controversial.
> It's so easy to see what everyone is working on, run reports to see how long things took
This, on the other hand is usually what's controversial, and the reasoning behind that is that it often leads to management using past data as a stick to drive "faster" future development, with silly statements like “Well, the average plugin took 2 days to write, why is this one taking 7 days? Oh it must be that the person doing it is lazy. Better give them negative feedback on it and tell them it can affect their promotions”.
The example is contrived, but the motivations and behaviors are very real. Competent developers tend to leave companies where they are constantly badgered about stuff like this by managers who have never developed software. In the extreme case (very few organizations are this bad, but they do exist), all you end up with is the remaining ones who know how to show that they are doing a lot of work, but are generally afraid to show their lack of knowledge in any area for fear of being dinged by the data-equipped management for it.
Since this is not conducive to the technical health of a software development team, estimation practices that purport to be unrealistically accurate are what's looked down upon.
Of course bad management is bad, but the idea that the type of metrics available to them make any material difference is not one that has any merit to my mind. A bad manager will always make your life hell, regardless of what measuring tools they have available.
Correct. That's why developers with bad managers strive to make as few measuring tools available to them as possible. In other words, if you're a manager and your employees seem to resent and fight against more ways of measuring "velocity" and the like, you should probably consider that your usage of this data and how you express the results might need some work.
We had a ticketing system and one guy in our group had between 60% and 80% more tickets than the rest of us. So it would seem we were all slackers, however he took the tickets that took like 2 minutes to do (password resets, account unlocks, etc...), meanwhile i would get stuck troubleshooting a production performance issue that would consume hours. Management didn't care what the ticket was, they liked to see pretty numbers of tickets completed.
I might be the only one who had this experience, and i am not saying ticketing systems are bad, they are extremely useful, as long as everyone knows how they work.
Honestly, I think the problem is structural with regard to management. Just changing the process or tool (ticketing vs constant communication with manager) isn't enough. It probably needs to be combined with smaller teams (i.e. Amazon's pizza rule) and management defining goals and "success" as more than meeting some narrow KPIs (MTTR on tickets, for example)
Sorry you had that experience, but to me the justification of those types of system are solid.
https://heeris.id.au/2013/this-is-why-you-shouldnt-interrupt...
Of course the right thing to do is to work to get proper visibility, and when you have that, then the "tap" is pointless interruption.
It is bad behavior, for two simple reasons:
1) The tapping of the shoulder (both literal and figurative) doesn't award any information, only the question afterward. So focus on what that question is, and implement whatever information that gets you (if any) into another system that doesn't require you huffing around interrupting the flow of your employees.
2) If that is your only method, you have already failed, and anything you do in that position that is not directly related to getting out of that position, is bad behavior.
This really hits the nail on the head, in my experience this is accurate for employers that are anti-remote, they want to be able to see people "work". I've argued that that means they are essentially paying people to be there 8 hours a day, not that they're working 8 hours a day. You see it a lot in stale offices, people are sitting at their desks appearing busy. Work doesn't get completed in the that time and they start hiring more busy drones.
It's absolutely baffling to me that we can't look at the statistics at a higher level and recognise that it's a highly inefficient way of working.
Ironically, many places will use JIRA/Trello/etc, and then still have the sync meetings where you repeat all the actions you've performed on the board!
We can look at them and your management can too. The important takeaway is not that they don't know or can't learn, it is that their sense of control over you is more important than any measurable truth. Once you realise it's just a dysfunctional personality trait it becomes a lot easier to understand why it's so hard to change.
So, home office for senior developers with awesome productivity only?
When I am working remotely, due to the organisation's inability to measure productivity, fostered by their management's "bums on seats" culture, I can't demonstrate that remote work is a benefit to them.
Why waste one person's time when you can waste ten?
David Graeber's book "Bullsh*t Jobs" is about that.
These companies main priority is not efficiency and productivity. If it was, they could hire less people to do the same work and that would foster unemployment.
They have a political and social role: hire as many people as possible. Keep them employed and busy and keep an eye on them.
Become too big to fail and you'll be rewarded with overbudgeted contracts (and even bailouts, sometimes).
Still with you here...
> They have a political and social role: hire as many people as possible.
You lose me here.
Most companies are out to make as much money as possible now or in the near future. I spent the first three years of my career in QA at a total BS Jobs company (120k+ employees, lots of government contract and enterprise software work). So many people coasted, nothing ever really got done, etc. Even in a totally dysfunctional organization like that, hiring was either about backfilling attrition losses or hiring people that were expected to directly or indirectly contribute to revenue growth.
Never attribute to a vast socio-political conspiracy that which can be explained by the emergent incompetence of a large organization.
The company in the abstract, sure. Not managers and VPs and such. Not even the C-suite, necessarily, or at least they likely have other concerns in addition to just the company making money. It's the principal agent problem plus the management version of résumé-driven-development, basically.
Even with 12 years of professional programming experience, I am only beginning to have an idea of how long things are going to take in the context of a small team. If you've never programmed before, I don't know how you could possibly come up with an idea of how long something should take. So it reduces to the heuristic of "well they're at their desk, I'm sure they're doing something."
They also save on office space.. I don't see how this is not a win win for everyone either.
Your ex CEO sounds like a psychopath.
Of course, one of the hardest professions to measure output for is software engineers.
I once worked with a guy who, it was discovered after a long time, actually had two Silicon Valley jobs. That he actually showed up at! He walked/drove between them a few times a day, and managed to keep the illusion up for quite a while.
I'm guessing he has 4 "remote" jobs now :)
the VCs aren't putting all their eggs in one basket with startups, so why should the workers?
I would suggest something like: "I've been crunching the numbers on how to increase efficiency and been pleasantly surprised at how well we're performing compared to our competition. I was thinking of trying an experiment based on this new study I've read <Link> where we leverage existing great management techniques to also allow remote work on a limited basis. If you're interested I could show you the numbers but I think with a small sample set <me and a few other good developers> we could prove that this could boost our efficiency even higher! In fact even though it's increasing our numbers we can even offer it to the team as a perk, <study> has shown that when a team gets a perk even if it boosts efficency they work even harder! We have this great new/old project <X> that would be perfect to measure the numbers with. How about I roll it out at the beginning of next month? Don't worry about the planning, I have some experience and can have a proposal on your desk by the end of the week."
For my part, I've been self-employed (full-time consulting), 98% remote, for about 3 years. Before that, in almost 20 years of traditional software-related jobs, about half my working days were remote. I was fortunate in being able to insist on a high degree of autonomy wrt how/where/when I got my work done, for most of my positions. Not always possible, but you'll never get the freedom if you don't look for the oppty and make the case for it.
It's then empirically obvious that I get more done and am happier.
If they don't like it, they can sack me. I earn more than twice minimum wage, ergo I can work half the year and get by.
It's never happened yet because it turns out actually people like people who get stuff done. Who knew?
Behind this monitor I'm looking out at the river, watching Winter roll in as I eat lunch. :)
working remotely gives too much power to workers because interviewing elsewhere becomes so inexpensive. In fact I'd argue most remote devs that are not already at the top of their potential compensation/status (say, someone working in the linux kernel team) are interviewing constantly. Most companies not aiming for top tallent are better off taking on contractors than remote full time developers.
Yeah, the answer to the OP articles question about why the remote studies work out that way isn't a big mystery.
The first thing you learn working from home is self-motivation and training yourself to get "in the zone". Then try doing the same at an office, where interruptions are far more often.
But that said not all offices are like this. A lot of companies work to 'protect' their creatives/engineers from disruption from the extroverts/talkers. Either by isolating them in separate closed office rooms or having rules like headphones on = don't bother them.
Well, yes, that's what you're being paid for.
The alternative is hard KPI's and getting fired for not meeting them, which is not something you really want.
Bill Gates: Measuring programming progress by lines of code is like measuring aircraft building progress by weight
And then, how do you measure someone refactoring code and ending up with -1kLOC?
Or someone thinking hard for a few days to come up with an elegant algorithm?
To help this movement, we built open source SaaS boilerplate. We also built Async, an asynchronous team communication tool designed specifically for small teams of software engineers.
It's kind of like how legalizing prostitution makes way more sense especially for the sex workers, but nobody ever talks of legalizing it because nobody wants to be the guy to bring it up or challenge the status quo.
Just leave and get a job at any remote-friendly company. The grass is really greener.
Management there only values "butt time in seats" and the ability to come over and interrupt you by tapping you on the shoulder whenever they need something.
As one of my ex CEOs put it: "If I don't see my employees stressing out at thier desks I get the impression they're not working."
Until we get over this psychological attachment of management loving to visually see their slaves on the open office plantation through their private panopticon[1] offices, remote won't take off no matter how many studies get published.
[1]https://en.m.wikipedia.org/wiki/Panopticon