Employers have increasingly turned to surveillance of keystrokes to make sure their workers are actually working, and this is especially true of people working from home. However, remote workers can fight back with hacks that fake keyboard strokes and mouse movements. The Wall Street Journal reports:
When Teramind examined an anonymized sample of 1,000,000 workers at 5,000 corporate clients late last year, it says it found 7% of employees appeared to be faking work activity on their machines. It has since fine-tuned the algorithms and rerun the test, this time finding over 8%.
“The true number is almost certainly higher,” Kleyman says, since the company found no false-positive results when it investigated the findings.
I've seen various measures of productivity for remote workers. Most suggest that it's lower, but some recent ones have pointed in the other direction. But I'd take this Journal article as evidence for the former. If 10-20% of remote workers are going to active lengths to fake work, that surely must mean less work is getting done.
For more on that score we can turn to the latest release of the American Time Use Study. Among full-time workers, it reports 8.19 hours of work per day for those in a workplace compared to 5.37 hours for those at home. That's a difference of two hours and 50 minutes.
Now, sure, that's not the end of the story. Some of those workplace hours were spent around the water cooler and some of the remote workers might have been hybrids. But there's no way that accounts for more than a bit of the difference. Nor is there much reason to think that the hours spent working are any more intense for home workers than for workplace workers.¹
The difference is just too big. Remote workers just don't work as much.
¹Though it's possible: fewer interruptions, less time on the phone, maybe fewer meetings, etc.
LOL No one spends time around water coolers. Talk about an anachronism. People spend their time shopping on the internet, keeping up to date on social media, or reading the news. They did that at work as much as they do when they WFH.
Exactly. And keyloggers hold bathroom breaks and reading memos or references against the person, but if you're in an office much easier to fake that as 'work'.
100%. People just waste time at work while watching the clock.
Number of keystrokes only matters if you write for a living.
Time thinking is time working. I walk away from my laptop regularly to think.
Anyone who's not an hourly rate worker spends way too much time thinking about work after work. We don't get paid for this.
I'm really fucking tired of being told I'm less productive working from home.
There is plenty of water cooler talk (well, coffee machine talk) at our office, as well as random drop-ins, spontaneous whiteboarding, and other chatter. While some of it is personal stuff and some office politics, around half is work related.
I find it sad that anyone thinks that keystrokes correlate to performance. In fact, the fewer I make, the better.
I think the difference here is in how people think about working. WFH people are answering about time the actually work during the day. In the office people basically count every hour at the office as them working - even if they aren't. I know some people that count the drive to and from work as work.
That's because it basically is.
And yep, bang on. When I was in the office, I "worked" 8 hours a day also, even though I'd tell everybody who asked that my actual time capital-w Working was less than half of that. Nothing's changed except that I no longer have to go through the charade of "working" for an large number of hours every week, simply to please our capitalist overlords and the work-as-religion culture they've created.
Definitely should count the hours commuting against your wages, otherwise you'll shortchange yourself and be salary rich but net income poor.
I have no idea whether remote workers spend less time working, but the largest impediments to productivity are, in my experience is meetings, workplace socializing, and web surfing, in that order.
If management made reducing the number meetings, keeping them on-topic, and ending them early or on time, the improvement in productivity would be astoundings.
Does the quality of work suffer?
I am on a hybrid set-up, Mon, Tue, Thurs. in office, Wed, Friday from home, and I work my ass off at home oftentimes right up to the last second of my day. My team loves that Friday is indeed the last day of the week and the next days are Sat/Sun, but we also dread Fridays as that is oftentimes the busiest day of the week for my team.
I work for Internet Brands, and lots if not almost all of their business units have a hybrid set-up, so there are more of us than you might think Kevin. Not to mention IB is not a huge company, but the total # of employees on their roster is in the thousands. So again, folks with a hybrid set-up seem fairly common post Covid.
Also yes, if I was in office and asked how often I am working, I would say the whole day, but also same when at home. Keystroke loggers are kind of a blunt instrument to log productivity and leave a lot of activity unlogged, time on the phone, taking notes by hand, actual scheduled meetings with boss/colleagues, last second meetings requested by Boss, etc..
I know you have a fascination with this topic, but man...dig deeper, it is easy to malign WFH/Hybrid workers in these articles you read, and as a WFH/hybrid for over 4 years now, I probably work harder than someone putting in an appearance in the office to look good to the suits.
My boss would most likely agree with me and have my back.
ETA: If the powers that be found that when working remote we not even 2/3rds as productive as in-office workers, they would force us to come in 5 days a week in a New York Minute, and it is telling that we have never been warned to shape up or we will ship out (or have to commute every day).
Also reading or looking stuff up - keystroke loggers completely ignore that input. Especially since many focus on the keystrokes within certain programs.
It also doesn’t measure time spent thinking.
I design training content, if I’m working from home and need to think about how to design a module for a course I might go for a walk and spend the whole time thinking about it, then get back to my desk and type it up. If I’m working in the office I’ll sit at my desk thinking about it then type it up.
Both take roughly the same amount of time, one is a whole lot healthier.
No, remote workers spend less time PRETENDING to work.
If (a) we have a lot more remote work now, but a similar number of total jobs as before we had more remote work, and (b) total productivity is on par or higher than before, then it MUST follow (in broad strokes) that remote work productivity is comparable to non-remote work productivity.
Oh lookee here, stats! https://tradingeconomics.com/united-states/productivity
(A) and (B) are true, in broad strokes, so therefore remote work productivity must be comparable.
In fact, look at that big bump in 2020. It sure looks like remote work might be MORE productive, given that a lot of non-remote work was temporarily dropped in the early days of the pandemic.
This post has been today's installment of "Kevin doesn't understand labor economics but wants to yell at cloud."
FOH with this stupid narrative, please.
If you're idling out or counting keystrokes, people are going to fake them.
Especially those who are efficient at doing their work.
This is a bullshit method of seeing productivity.
Even more, I"d bet that with 5000 workplaces using keyloggers, far more than 7% of them are using that keylogger to idle workers out of the remote software. At my spouse's last workplace, one of the companies she worked with had an invisible timer - much like the edit timers here - on the input forms. So if a form was particularly log, these was a high chance she'd have to re-enter everything all over again.
It's just executives making dumb decisions.
It's always funny when Kevin "that's not a big deal" Drum gets a burr in his hide about something, and for the last few years one of the more consistent somethings has been "remote work bad".
Especially since he's remote-worked exclusively for over two decades, and apparently doesn't think any of this applies to him.
Sorry, Kevin, I think this survey falls under WhoGAS.
For WFH, it's output -- not time in seat -- that should be measured.
Are remote workers making as many widgets as their in-office peers? If in-office workers produced a third more widgets (as would be expected based upon the extra hours "worked") that it should be pretty obvious.
i once got a marginal score on a corporate review; the number of features i implemented was fine, but got dinged for fixing fewer bugs than others on my team
mainly because i spent a lot of time writing test code and fixing issues before putting stuff into production
but i learned my lesson; i still wrote my test code, but i didn't actually fix minor things until someone pointed out the problem in production and they were put into the tracking system
my next review i got better marks and an attaboy for fixing my bugs really fast
of course i left that job shortly after
I here you on that one. I was once asked by a team leader how I was writing the most productive code even though as far as they could see I spent most of my time fiddling about on yellow pads. Rolls eyes. No, counting the number of keystrokes in an hour is seriously _not_ a good measure.
i worked at one place during the internet bubble where the CEO forbade whiteboards
i shit you not.
You've got to be kidding me. Seriously. You've got to be kidding me. What was the justification? I don't think I'm the only one here who would like very much to know why.
he thought they made the office look messy
also, winter boots/coats were not allowed in your office, had to be in a closet (so no keys, wallets, etc in your coat)
purses had to remain in desk drawers
no family photos
the only thing allowed on your desk when you left for the day was a monitor, mouse, and keyboard; chair had to be pushed in, not away from the desk
Some of it comes from the olden days when AT&T decreed that 10 lines of debugged code/day was expected productivity. Maybe it made some sense in the day of one-off projects and non-reusable code.
Let's consider two cases. Given a set of requirements:
1. Software developer implements it in two weeks writing 120 lines of code (exceeding the AT&T standard).
2. Person implements it in two weeks writing 90 lines of code (slightly below AT&T productivity metric).
What's more productive? Certainly measuring keystrokes, person 2 gets fired; person 1 gets promoted. This is where we are today.
This is why I roll my eyes and never tire to point out that software development productivity metrics are complete and utter bunk.
+1
Open access article in Nature, albeit chinese company:
https://www.nature.com/articles/s41586-024-07500-2
As a result of this research, the whole the company went part WFH.
"When your metric becomes a target it ceases to be a measure", or so goes the saying. What evidence does Kevin have that keystroke loggers are a good proxy for productivity, and what evidence does Kevin have that Teramind (God, what an awful name, who makes these decisions) does good statistics, or indeed, is not motivated in the slightest to 'find' the things they've been contracted to search for. The whole thing smacks of 'expert testimony' by people who bill themselves as 'expert lie detectors', or companies that claim they can find what text has been written by an LLM vs an actual worker or student.
Also: Real estate values!
This reminds me a of a great story Roger Ebert told about writing Beyond the Valley of the Dolls. He was set up in an office adjacent to Russ Meyers, and Meyers equated writing with typing, if he didn’t hear the typewriter he’d yell through the wall for Ebert to get back to work.
That dovetails pretty well with David Graeber's book, Bullshit Jobs, which describes very many white collar jobs.
just wondering if kevin thinks he'd produce more/better posts if he had to spend 3 hours a day commuting to an office
office space is expensive in socal, but he can do a cheap approximation by just sitting in his car in the driveway for 1.5 hours at 7:30am then again at 5:00pm
(and no cheating with cell or wifi!)
> If 10-20% of remote workers are going to active lengths to fake work, that surely must mean less work is getting done.
Not really. Back in the '90s, I had an office job but would work at home maybe once or twice a month. I always said, and it was true, that I got as much done in 4-6 hours at home as I did in 8 hours in the office. No socializing, no interruptions with things that didn't really need immediate attention, no phone calls... Of course, I didn't have to justify how I spent my time at home except by whether I finished my work. If someone was somehow actually tracking my hours "working," damn straight I'd fill those extra 2 hours with fake work.
Yeah, and in general if employers implement infantilizing surveillance policies then employees will react accordingly and work-to-rule to some extent. If the rule is that I need to show up as active in chat during work hours then you can be sure I will show up as active in chat, and if ensuring that I do messes up with my flow or gets in the way of actually getting tasks done in an efficient way then so be it.
It's the old adage that "when a measure becomes a target, it ceases to be a good measure". In practice employers demand employees to appear to be working, so that's what they get.
Why bother? Sprinkle treats on your keybaord. Your cat will do the work for you.
For Christsakes Kevin stop referring to the American Time Use survey, it’s been pointed out to you repeatedly that it measures all work from home not just remote workers. As for Terrormind there’s zero proof 7% of workers are faking work they’re trying to sell something to scare companies. If productivity of WFH workers suddenly declined by almost 10% the whole economy would know not just the company. The fact is Kevin hates remote work for some unknown reason and can’t be relied on to discuss the issue impartially.
I have had the same job since before Covid. I could work from home but also needed to be onsite. You can’t judge my productivity based solely on where I was located.
There's a really basic flaw in counting keystrokes. Keystrokes are input; the value of a person's work depends on their output. The relation between input in keystrokes and actual output is only statistical. Some workers are very accurate typists and can get keyboard work done with a minimum of keystrokes. Others (me, for instance) are not very good typists and spend quite a lot of keystrokes making and correcting errors. For a given output, I probably input a third more keystrokes than a really skilled typist; does that make me a third more productive?
Kevin is rightfully getting lambasted for this one. I'm full time WFM and my primary task is help desk support/training - so there's a significant chunk of time spent talking and watching the end user via teams/skype to tell them what they were doing wrong. Also I use my down time to troubleshoot via running exception reports, throwing it all in a database and seeing what's wrong. You can't measure all of it via keystrokes. I self report to my boss each end of day what I did and also log help tickets into our team site.
This immediately raises the question of why they didn't report what percentage of workers are faking activity in the office. That omission suggests that it is about the same as for home workers.
You would imagine that companies hire people to do things and then evaluate their performance by measuring what they get done. That would provide a reasonable way to evaluate whether office or home workers get more done. Instead, we get talk about office morale and curious metrics about keystrokes per minute or whatever.
If businesses exist to maximize shareholder profits, then the best metric would be whether corporate profits and stock price are up or down when comparing competing companies with different mixes of office and home workers.