Former Microsoft CEO Steve Ballmer on useless performance metrics[1][2]:
>> In IBM there's a religion in software that says you have to count K-LOCs, and a K-LOC is a thousand line of code. How big a project is it? Oh, it's sort of a 10K-LOC project. This is a 20K-LOCer. And this is 5OK-LOCs. And IBM wanted to sort of make it the religion about how we got paid. How much money we made off OS 2, how much they did. How many K-LOCs did you do? And we kept trying to convince them - hey, if we have - a developer's got a good idea and he can get something done in 4K-LOCs instead of 20K-LOCs, should we make less money? Because he's made something smaller and faster, less KLOC. K-LOCs, K-LOCs, that's the methodology. Ugh anyway, that always makes my back just crinkle up at the thought of the whole thing.
Maybe the question of "How do I measure developer productivity?" is to broad to be useful?
Per the article, what managers often actually want to know is "How do I detect when a developer is so unproductive that they should not be retained?".
What are some behaviors that developers have when they've "checked out"? Not showing up to work or logging in is an obvious example. Lines of code might be a poor metric, but maybe the bottom 1% of coders in that metric is a sufficient detection for poor performance?
...at least as a detection mechanism to highlight for further investigation?
The manager rarely needs to detect when the developer is that unproductive - it's glaringly obvious. In practice, even if you as the manager don't notice, the team will tell you that another developer isn't pulling their weight. However, you can't just fire someone in most organizations because their teammates have a gut feeling -- you need to have evidence as you get a case together. Further, some causes of poor performance are temporary; some are an indication of a lack of skill that can be remedied; some are that someone is a good worker in the wrong place in the organization; Sometimes, a developer is just lazy. A good manager will be able to discern this and determine if the solution is to change the environment in a way that brings success to the dev.
But if the case is that the developer needs to be "managed out," you need to build the case as a manager. This is where those metrics help.
The first Turkish software export was measured in Meters :)
The story goes, a Turkish company sells a software to a British client for a few hundreds thousands GBP and has to go though the bureaucracy for the accounting purposes.
Turns out, the clerk in the customs doesn't understand how software works and says he cannot certify the export of a few hundred grands worth of goods delivered by press of a button(the delivery was done through 28kbps modem connection). They bring him a disk but this is not good enough, there is no way this piece is worth that much he says and rejects the application.
They end up putting the software on a tape and declare that they've sold 2000 meters of fine Turkish software to the Brits. The clerk likes that, so the first software export from Turkey ends up being measured in Meters!
[0] The sources are in Turkish but the guy who was involved in this is Ali Akurgal.
I can't find a reference now, but there's an old story about an early import of some software from the UK into Ireland (in the 50s or so). The software came on punched cards, and when it arrived turned out not to load. After much investigation it turned out that some cards were missing, so the vendor sent it again. This time different cards were missing. Eventually the problem was tracked down; at the time, it was normal for customs officials to retain a small sample of any bulk material imported for post-hoc inspection. The punched cards were viewed as a bulk material...
Also the same story highlights the roots of the term "Tape Out".
Since the designs of ICs were stored in the tapes at the time, you write the final design to a tape and send it out to fab, hence you Tape out the design.
I recall from some HN story that "tape out" is an even older term, referring to the practice of putting black line tape on the (much magnified) artwork for the photomask. Wikipedia seems to confirm this: https://en.wikipedia.org/wiki/Tape-out#History.
Twenty years ago I worked for a company that had in the past paid their contractors by LOC count.
Turns out if you base the pay on that ... why create a function and call it repeatedly when you can just copy and paste the same code block over and over again
If you know a thing or two about evolution, it should be obvious how these shortcuts will play out.
And still, you hear about similar metrics being used all the time... All over the place.
I suspect, in most cases, it's actually bad metrics stacked and really the manager trying to putting a metric on their own "contribution". (Parasitic isn't even the word, but rather viral or prionic...)
In the end, truly and reproducibly recognizing individual contribution is probably beyond human comprehension for any non-trivial tasks. Maybe your product actually benefits a lot from that one guy, who is just very good at promoting a good spirit, being fun to work with, rather than being the most proliferative coder. And how do you measure how hard a problem is? Does is matter who perceives it as hard in your team?
I think, it's one of those things, where trying to be clever will most likely make things much worse because the manual labor and personal involvement of the past actually was already best fitted for the task, by utilizing the human brain where it excels. Humans doing human things with humans, without being willfully ignorant about the complexity at hand.
It's always fun trying to explain to non-tech management the two truths of managing software development:
1. There is no objective measure of productivity that can be applied to software development.
2. Accurate estimation of development times is impossible (not difficult, actually theoretically impossible). All estimates of development times are wrong, some by more than an order of magnitude.
The second one is especially hard to grok for non-techies. I have to explain that if you insist on accurate estimates, you will get grossly padded estimates [0]. And the work will expand to fit the time available (sometimes resulting in the task exceeding even the massively padded estimates because the work expanded with the estimate). I've had many entertaining conversations attempting to explain this.
[0] Because every developer knows that an estimate will magically become a deadline.
> 2. Accurate estimation of development times is impossible (not difficult, actually theoretically impossible). All estimates of development times are wrong, some by more than an order of magnitude.
Personally now-a-days I care more about velocity. If the team can deliver the smallest value added to the system within a week, we are good. The more constant velocity the team has, the more Management trusts you to produce value and get their things done without the need to have massive estimation parties that almost always end up being wrong.
The sad thing is, once the velocity gets bogged up or waving (Org change, contantly changing directions,..) the estimation parties are here to stay and it's quite hard to get back to where the team was.
> 2. Accurate estimation of development times is impossible (not difficult, actually theoretically impossible). All estimates of development times are wrong, some by more than an order of magnitude.
Some are useful. The goal isn’t to be right in that there’s no award for perfect estimates, that would be stupid. But having an estimate, especially with relevance to multiple features is helpful.
At one time a team I worked on used “story point cards” and each person would estimate blindly and the discuss. It was interesting hearing people’s reasons behind their magnitudes.
Over time the story estimates got pretty good. But the number was completely useless objectively and made no sense when comparing teams or different projects.
> The second one is especially hard to grok for non-techies. I have to explain that if you insist on accurate estimates, you will get grossly padded estimates [0]. And the work will expand to fit the time available (sometimes resulting in the task exceeding even the massively padded estimates because the work expanded with the estimate). I've had many entertaining conversations attempting to explain this.
Sounds like Hofstadter's law.
"Hofstadter's Law: It always takes longer than you expect, even when you take into account Hofstadter's Law."
Now think about money as a metric for human prosperity. Which creatures really roam the markets and where is mankind's place there? Is any dynamic stability of intrinsic value to us? (Not trying to be edgy, or deep, I just think those are the questions indeed usually not answered by free market enthusiasts.)
If you give a bug to a software developer to fix, the good software developer will probably rewrite some code and in the end the LOC count will be the same or less. The bad/lazy software developer will glue some fix on top of the bug and add additional LOCs.
Depends. Sometimes software is so convoluted that you're sparing yourself and your colleagues a lot of trouble by adding that one liner rather than doing a big refactoring, especially in the absence of tests.
KLOC can be a useful productivity metric to track at an individual developer level, given the developer works in the same domain for an extended period of time, and has the professional and skill level maturity not to game it. That relies on management that doesn't try to aggregate it or compare it in an apples to oranges fashion across developers, domains or (worse) languages.
The problem is that this is a theoretical scenario which rarely comes up in reality. The developers operating at the skill and maturity level where tracking their individual productivity and quality can be done well, generally are not the developers who really need these types of measures, and they usually aren't left grinding on the same domain for the years required to have enough data to do anything useful with it.
It is kind of a catch 22. The situations where tracking and gaining insight from KLOCs would help (mature/skilled devs working on the same thing over and over for years...doesn't happen much), are not the situations where it is most needed (projects with hundreds of relatively inexperienced/immature developers and managers)
The result is it is applied to disastrous effect in situations where immature managers glom onto it as a silver bullet for managing the unmanageable.
I’d rather have an employee that spent three days, optimizing a single 100 LoC method, than one that churned out 2,000 LoC that does the same thing, in two days, that will need to be maintained, forevermore.
The less code the better! Usually when I see massive PRs with hundreds of lines of code, the engineer who wrote it has through lack of experience or lack of humility re-written a huge chunk of stuff that's already in the standard library.
Less code reaches a limit where readability and future maintenance costs more. The art is having just enough code to do the job without having too much or too little. Unfortunately, programmers are not often measured on their ability to write code that can be handed off to others and maintained for the next 20 years.
Was Jobs really that smart? Not sure that's the niche he filled successfully. More like Kinski, talented in some domain... but smart?! Jobs died pretty much by Dunning-Kruger effect, no?
People can be very intelligent on one area, and not as much in others.
There are many scientists who have very high standards for proof and evidence in their professional life, but accept other theories as truth without a shred of evidence in their personal lives.
Sure, but I think it's also easy to conflate success and talent with intelligence. I think many successful scientist are not particularly smart, but rather dilligent and educated; most innovation is not a leap, but a steady incremental progress which may accumulate in a "innovative" product at some point.
I don't know much about Jobs, that's why I asked. From what I got, he was rather a charismatic leader with a good intuition, but I wouldn't call those attributes intelligence per se. Is there any evidence for a particular intelligence of his?
>These metrics include how often workers turn their cameras on during virtual meetings, how frequently they send emails (and how many contain @ mentions), whether they regularly contribute to shared documents or group chats, and the number of days they used Microsoft’s tools such as Word, Excel, Skype, Outlook, or Teams in the last month
The worst part is that none of these are results-based metrics. These are precisely the kind of things you would measure if you want to create lots of extra meetings and redundant communication.
It's also incredibly self-serving of MS. They have essentially defined "productivity" as how much time you spend using MS products.
I hope natural selection leads to the shuttering of all the companies which actually try to use this metric to optimize their organization.
Yeah, that's what I thought. This is a measure of "how well you're using Microsoft software" so that Microsoft itself can use it to market more software and services to you.
I really like @mentions in email. They help me find specific stuff relevant to me. And also makes it a second faster to include someone in the to line.
I wish they displayed better because in my org they show as “Prepend, Prependson (ABCD/YTD/Foo/Bar) (Contractor/Non-Contractor Status)” so it’s stupid looking in the body.
I’d hate that so much. People are already abusing CC and BCC in corporate email chains, having to simply @ someone to send them a notification or a copy of the email would make that problem even worse. MS always comes up with things that make working so much harder on the business/office/windows side.
Yeah I feel like it's also adding a redundant extension to email, which is a perfectly good and time-tested technology. It's just one more way to create a wedge feature to create an inferior experience for people not using the MS product without actually adding any value over what's already available from CC/BCC
It would be interesting to see if I include more people with this feature. I think that I use it only when I would have gone up and added them to the CC line, but I would have to try to call out what I want them to see in the email.
@FalconSensei Want to get on Teams and collaborate to fix the spelling mistakes I accidentally-on-purpose left in all those Office docs?
@friend1 @friend2 @friend3 Don't forget beers on Friday.
Lol.
This is so stupid. You'll end up with 2 types of employees; those that work and those that work to game the metrics.
My mother's seen it first hand at a government job. Shared work queues and measuring productivity via naïve metrics is an ideal work environment for lazy people that do poor quality work. Why spend 30 minutes on a task when you can do a terrible job of it in 5 minutes and get the same score?
Thing is, in her department she was the only employee that took difficult cases and found the solution, and left the easy ones to her coworkers, her boss didn't want to fire her, but had to because the metrics said she had to go because a low total amount of cases...
Last she talked with her former coworkers, now they are struggling because the hard cases keep piling up endlessy and nobody knows how to do many of them, because they fired her without warning so she couldn't even teach others how to do what she did.
EDIT: oh yeah, remembered that IBM wasn't subtle about how much they needed her either, since she was the only one that could do what she did, when she worked at IBM I had lovely christmas and new year parties with her laptop on the table, because IBM back then threatened to fire her on the spot if she didn't work remotely on some important clients during those times, then her actual firing came some a short time later anyway during a seasonal lull of work.
My first job was in tech support and we had similar metrics although nothing as drastic. I took really hard escalations and had a low case count. My manager pointed it out and so next period I also did 100 remote password changes in addition to my hard cases and had like 3x the next closest case count.
I did this to show him how stupid his metrics are but he just said “great job, this is amazing.”
So I just made sure to min max stupid metrics and do what I thought was right as it was easier than trying to fix the metrics.
I definitely don’t think it’s worth being fired out of principle if I know I’m measured for case count.
Generally, I try to set up policies where these metrics are not allowed for performance, but are visible.
And then whenever someone tries to enact them I try to say “this is a really stupid idea and here’s why” in front of as many senior, fancy pants people as possible.
This is why metrics should all be taken with a grain of salt and be subject to the opinion of ones manager. Nepotism at level certainly exists, but a good manager will always come to bat for you. Even if it sticks their neck out. That is why upper level bourgeoisie management that has no f'ing clue about the bottom rungs shouldnt touch what their not ready to unravel.
So frustrating to read that. Could no one even look at 'type' of cases? Tickets in a ticketing system typically have some sort of severity level associated. Taking 3 'critical' issues should be treated as at least important as taking 30 'low priority' issues.
I had similar experiences in the past with review comments. So, they measure how many merge requests you left a comment (for other people's MRs), how many comments total on other peoples' merge requests, how many on your own.
So I have seem approved MRs with 1 comment: 'lgtm' (looks good to me)
I have seen the exact same comment pasted in all instances of something, in the same MR. i.e: "don't use @autobind, remove." instead of 1 comment saying 'remove all @autobind from file'
Situation above, the person replying: 'fixed' on all comments, instead of 1 reply 'removed all @autobind'
Opposite anecdote - worked at a place where I was ... 6 feet from the main sysadmin guy. I would routinely email him, both as a bit of CYA, but also because his schedule didn't jive with mine. Owner would come out sometimes and get irate that we were emailing vs "just talk to each other - you're sitting right here!". But... I didn't want to interrupt the other guy. In fact, I got better results by emailing - more details, could follow up as needed, and he would respond with 'done' so there was an actual log of the interactions. When I'd just lean over and interrupt him, everything was slower and worse. The '2 minute' things rarely were 2 minutes, and would take away time from what he was doing. If we met in the lunchroom... yeah, I might say "hey, I'm going to need X"... and he'd say "email the details".
If that guy is anything like me, he probably treats emails as tasks. Each email is something that needs to get done. Now I may be doing something else so I need to prioritize that over your thing right now. Overall I think that's why it worked out so well for you guys.
I see your point and I agree with that, as long as it's not a metric, but the way you chose to communicate because it was best for the situation.
I've been on both sides of both situations (in person better, online better). But what I always did was: ok, feel free to interrupt me if something urgent comes up, otherwise, send me a slack with the full details and I'll answer later.
As mentioned in other responses to this post, having the number of emails/slack messages as a target doesn't help make the work environment better (in both 'more pleasant' and 'more productive' ways)
Isn't it essentially a good measure then, just accidentally flipped around? Just flip it around and it's good.
A truly bad measure would be something like measuring the volume of coffee in my cup to determine the health of the economy. If I make up fake numbers, or pour some coffee out to change the measurements, that doesn't make my measurements any more insightful in regard to the economy.
>Nice, I guess instead of talking to the person next to me and sorting something out in 2 minutes, we can use email, and do it in 30.
I do agree and disagree.
Sometimes it's very handy to just talk about stuff quickly, but for me, who unfortunely has a lot of context switching I love having stuff written in email, because I can always go back to it when I forget or when I need to reanalyze it
I see your point and I agree. I hated being constantly interrupted on my previous job, and I got them to stop doing that and only interrupt me when something really needed immediate attention.
But the thing is: communicate in the way makes sense for the situation/environment. Not by forcing people to send everything in emails, and even, sending way more emails than necessary
I turned off email notifications and check it only in the morning, or if I am specifically waiting on something like an access code. Oftentimes, something "urgent" that was sent at 13:30 ends up with a "never mind, I found it" response at 15:00 so I can just delete both emails.
>Nice, I guess instead of talking to the person next to me and sorting something out in 2 minutes, we can use email, and do it in 30.
Well, talking to the person next to you will disrupt their focus, and they might take hours to get into whatever they were doing (if they were programming / in the zone) at the time. So, yeah, better do it in 30 minutes.
That requires several context switches on both ends. I'd rather have my focus interrupted once than having to either draw out an email conversation over several days or mentally switch to it every few minutes.
In what universe do you currently talk to the person next to you? In an office job that would be horrifically irresponsible. I agree with your broader point, this is just a poorly chosen example.
"Currently" being emphasized I guess you are thinking about COVID and social distanciation?
Nonetheless, it’s quite common to read comments about talk in the office as a focus annoyance, flow interruption, etc. I guess any form of notification is.
I’m more wondering where it would be. I work in the public sector, so we almost literally employ people from every field imaginable aside from maybe academic research, and aside from the camera thing, I can’t think of a single one of those metrics that’s positive.
Maybe Denmark is just different, but even places like finances that still use a lot of excel, they also use dedicated stat tools, we use one called Targit, but PowerBI or whatever Microsoft’s version is called these days would be similar. Even people who do a lot of writing, like our communications staff, would only use word for drafts and maybe spell checking (if at all) before they move their writings into their specialised systems.
This is not to say that we don’t use a lot of office, we do, but it’s almost never a primary tool aside from collaborating, and while logging more camera hours may be desirable (for us anyway) sending a lot of emails isn’t necessarily a good thing, mostly it wouldn’t be.
On the flip side, and now that I’ve talked myself into it, I actually do know of managers who would value this sort of logging, shit managers but managers, so maybe Microsoft is just producing what their clients want? Anyway, if that’s the case, they probably shouldn’t add support for bad management practices. I know they have no real competition to office365 for non-tech enterprise, like at all, but who knows when that will change with Amazon around smelling all those billions.
Most commonly seen as part of login messages, as many Unix and Linux systems come or are configured to pipe a fortune message along with some other fixed text.
One could score very highly in these metrics without actually getting an ounce of work done. The inverse is also true. Scary.
This is wholly bizarre because these are in no way productivity measures. These are measures of software usage, which is being (wrongly) used as an correlation for productivity.
I wonder what the real motive is, because surely no one who implemented this actually believes it measures productivity?
And it doesn't measure normal phone calls or on site support for customers. There is nothing better than having to run from one issue to the next and getting a mail saying I just enjoyed a week of peace and quiet.
There aren't that many jobs where white-collar workers aren't forced to use Office and Outlook, and that's not just because of lock-in, but because there's literally no other alternative that doesn't suck badly. And no, Google office suite doesn't count; it barely matured out of toy stage recently.
MS tools have also huge, compounding advantages for larger companies who go all-in - they interoperate deeply. UX-wise, it's all quite not bad, and I could even bear it being proprietary; I just wish they didn't put all that telemetry crap into it - MS is ruining both their OS and their office suites by doing that.
Any business that may potentially compete with Microsoft (or Google) would be pretty foolish to use their non-e2e cloud services to go about doing so.
We've seen how that ends for Amazon sharecroppers.
I'm also not 100% sure about your main claim regarding white collar jobs. There may be a majority of positions that use the MS ecosystem, but there are still very very many that don't, even if they are marketshare-wise only 10%.
Really, I think it's mostly just Excel, and maybe Sharepoint. You're absolutely right that there is no great Excel competitor.
Yeah, I meant primarily Excel and Sharepoint, and maybe MS Project (though that has commercial alternatives). But also that once a company commits to MS ecosystem - including Sharepoint, Outlook, Teams, OneDrive and the whole Office suite - all that software suddenly gains a lot of additional functionality thanks to deep interoperability.
> Productivity Score identifies areas where you can offer people training to learn how to use the tools to their fullest capacity
In other words, it's a score that tells the Microsoft sales rep where to target their sales. If you score highly on everything, they'll give you a pat on the back to tell you that you're amongst their most "sophisticated" users.
This seems common in the enterprise sales world. Another example I've seen is Adobe, which holds an annual Symposium every year where they hand out awards to the most advanced users of their products. The "winners" get to feel like they've been recognized for their incredible digital marketing skills and the rest of their clients are shown how to increase the amount of vendor lock-in and dependence on Adobe. Everyone wins!
> This reads to me as “identifies people that should be managed out”
That's a very pessimistic outlook I think.
I work as a consultant for a company with approx 100 employees. Average age of workers is somewhere around 45 and most aren't really that good with computers, decent at best.
I've handheld their digital security officer when he configures a Zoom webinar. So far, he hasn't been sure enough to do it all alone and still asks me to check the settings.
This company produces word documents as their core business and if their digital security representative is that bad with computers, you can probably guess the average skill of the rest of the company.
Nobody is getting fired though, they still get the job done and management invests in training. I'm currently helping build their entire IT ecosystem and one key aspect of the tools selected, is that it's simple enough to teach to their employees. I'm sure they'll love those metrics as they can actually see the biggest pain points and proactively do something about it.
Microsoft tries to soften the pill with positive vibes, but they literally named their product "productivity score". As a manager, I would expect a "productivity score" to reflect the productivity of my employees. If someone has a low productivity what should I do? Of course I can provide training. I should also avoid raising his salary. And when I have to fire people, I will fire people with the lowest productivity.
I don't think this is a cynical or pessimist point of view. It is exactly what Microsoft is selling here.
Incompetent in Zoom, super competent in digital security.
Lots of people are incompetent in things that aren’t important. Feynman famously said he was incompetent in administration [0] to stop people from assigning admin work.
Competence isn’t a universal trait where people are perfect in everything or stupid in everything. I don’t get mad at a wizard because he’s incompetent at full plate.
I don’t think that’s true. It depends on what you’re using zoom for and how.
Tons of security conscious people use zoom, and use it appropriately. I assume any firm big enough to have a digital safety officer is going to have some large user base needing and using zoom.
> It depends on what you’re using zoom for and how.
No, it really doesn't. Zoom has had so many incidents at this point where the client would compromise endpoint security that the only reasonable approach is to consider any computer that has ever had the client installed to be infected.
Yes, there are precautions you can take to reduce the risk (burner VMs, and so on), but as a security professional you have a responsibility to lead by example. You yourself might be safe, but you're effectively telling everyone that Zoom is an acceptable platform to use, and that's just completely irresponsible.
Actually, my impression is that microsoft software targets IT more than the employees. The more power and convenience they give IT, the more likely they will get sales, because... who picks the tools and approves the sales?
Not that great for the employees though. As tools they are very meh. I'm surprised people are able to get anything done.
> The more power and convenience they give IT, the more likely they will get sales, because... who picks the tools and approves the sales?
I don't believe I've ever worked in an organization where IT chose the tools and approved the sales. It is, in my experience, quite common to have to code around a lot of shitty tools who's purchasing decision began at some executive's golf outing.
Slack has been doing similar things via their Analytics for several years now and their Customer Success teams regularly meet to discuss “Deployment Maturity” with larger customers on Enterprise Grid. In these PowerPoints you get their take on where your organization is at in employees engaging with Slack, compared to some of their other bigger customers, and they drill into how many applications you’ve got deployed, how many custom ones you’ve created, whether you use features like Workflows, etc.
There are statistics for what % of employees are using Slack, how many days in the month, % of messages being shared in public channels, private channels, direct messages, % of messages sent vs. % which actually get read, ...
What Microsoft is doing here seems like they’re trying to copy what Slack has done, to try and drive ROI and renewal conversations with their customers.
It isn't used as part of management or anything so it is just a fun little fact, but I was surprised to learn that such a number was readily available. It shouldn't have been surprising, but it was when it clicked in my head that Slack tracks you as much as any other app.
It's put next to "# of AltaMLers who think strawberry is the best Bubly flavour" and "% of AltaMLers who think pineapple is an acceptable pizza topping", I don't think it's supposed to be taken super seriously
Those measures are not that useful at all without a denominator. Do you have 10,000 employees or 2? 2880 slack messages means nothing without knowing the denominator. Same for countries of origin.
I wonder who sees these and thinks anything other than “what kind of dipshits think this matters?” Although, I must think that a lot of the people I know of who are buying ML consultants would likely think that counting and displaying random metrics without context is data science. So maybe it works well.
"Oh someone else does it, then it must be fine". I wish people stood up against such invasive practices and maybe we should have a legislation regulating this if company greed clouds being decent.
Most of the time these days MS news is "cool." Open this, Linux that, check out our cool terminal. Nice to see a reminder they are still definitely not cool.
MS is certainly trying to play the new cool, but the DNA remains the same DNA from the '80s & '90s. Don't be fooled: A Leopard cannot change its shorts.
It literally doesn't. All the leadership from that time has retired, the revenue sources are different, and even the business models are different (subscription vs. upfront for many services).
> Don't be fooled: A Leopard cannot change its shorts.
Someone should tell this to all the large companies started by or tied to Nazis (Porsche, IBM, Volkswagen, Puma, Adidas, etc.)
Or they should tell Apple, which went from near-death as a desktop/laptop vendor to a company now making most of its money from services.
Companies change. They do a mixture of good and bad things. At this point, Microsoft is far, far from the company they were in the 80s and 90s, even if they still do bad things.
That's revenue. Profit is more important to me. An extreme example is when game consoles are sold near (or even below) cost in order to create a user base for game sales.
Apple's services have at least twice the margin of their hardware, as far as I've seen[1].
> Microsoft 365's corporate VP Jared Spataro specified in a blog post that the feature, which debuted to little fanfare on Nov. 17, is “not a work monitoring tool”
Hahahahaha. A comedian of that caliber should have a Netflix special. That's like the warning on the Q-Tip box that tells you not to stick them in your ears. Of course it's a monitoring tool. They just don't want to call it that because it'll damage the brand.
> These metrics include ... whether they regularly contribute to shared documents or group chats, and the number of days they used Microsoft’s tools such as Word, Excel, Skype, Outlook, or Teams in the last month
Workers are incentivised to maximise their use of Microsoft products, and in turn the high employee utilisation of Microsoft products ensures that Microsoft continues to sell licenses and subscriptions.
This feature seems to me to be a real mistake from Microsoft. From a PR standpoint it's terrible, and it's a tiny feature that I'm sure drives very few sales.
What does microsoft care about sales? This isn't the sort of thing that would make someone used an MS product over an Apple or F/OSS product. Such decisions are generational, occurring only once or twice the life of a company. So at the end of the day this is a "feature" being added into a product that people are already using. Nobody really chooses to have it or not. It just becomes an option that will sit quietly in every office machine or remote access account just waiting for an evil manager to flick the switch.
Hey Microsoft! Can I pay extra for fork of windows/o365/office/whatever that doesn't and will NEVER have this ridiculous feature? I am putting my marker down. I am willing to pay MORE money for LESS evil in my machine. The more evil ideas you come up with, the more I will be willing to pay to keep them away. That is the reality in which we live.
I’m pretty sure this would be illegal to enable in many places. At least in Sweden the law explicitly prohibits monitoring of work performance in computer systems.
This is the kind of thing that is used to stop unionisation and in most cases to check if there's a serious productivity bottleneck, it is probably more valuable to microsoft than sales in the long run.
How come? In many countries it is enforced by law, and in some of them, everyone on the building gets it, regardless if they are cleaning bathrooms, or doing CI/CD pipelines on a random cloud deployment.
As bad as it is, isn't it great that they announce that they are spying on us? They wouldn't really need to - how would we know, if they didn't?
Enough naivety...
If people don't think that this has been going on for years, or that they actually announce Orwellian surveillance, I'm not sure what to tell you! We have been already living in a track and trace world - what is happening is that it is becoming overt. And there is nothing to be done about it. We all need the paycheck, right?
Even if there is a massive outcry about this - and I doubt there will be - they will just come back and do it another way. They claim it is a right for employers to get the 'metrics'.
What's kind of dark, twisted and funny, is that the people building the cages of today and tomorrow, are doing so with open eyes - technologists are some of the most surveilled people about, but - no problem! - we'll code tomorrow's hell for you!
There's a difference between hearing that such things happen, and suddenly finding an "Insights" (or more recently, "Cortana") e-mail in your inbox, telling you how much time you've spent on meetings, and "collaborating via e-mail and Teams", over the past week, complete with a friendly question like "did you make good use of your focus time this week?". The latter can really make you viscerally aware of just how much you're being surveilled, and in concrete terms.
(Or the helpful Cortana e-mailing you excerpts from your conversations, in which it highlighted what it interprets to be pending tasks. And because it comes as an e-mail, you know it's not being computed locally.)
What bothers me is that this system probably works at M$, which means their managers are probably using it to monitor their staff, using their inhouse tools. That probably has a good ROI since finding issues with their products is helpful.
However, for another company this metric is just a value of how well locked into the M$ stack they are. If it were my company, heavy users would be penalised as there are other, better, less expensive products to use. Sitting in Teams talking to people is not a good score. Firstly you're wasting 512MB of that company laptop on a single web page, secondly it is a slow communication tool and thirdly, if you were talking in IRC then the conversation would be logged for others to grep and find answers to similar issues.
Scoring high with M$ Word would also be a negative thing in my view, that shows that you're using a tool that locks the company into an upgrade cycle for future M$ Word versions. Did you not think to use Libre Office or perhaps MD?
Scoring high in Outlook is also negative, that's another tool that can't store .msg files in open way. Every time a .msg file is attached to something and exchanged, via another tool then the receiver needs something that can disassemble the .msg into a readable form.
Why stop there? Why not define networks of employees and their relationship? Why not weight an employee productivity score by the productivity score of their contacts? The possibilities are endless!
I worked for a startup that was deploying a system like this to a large fortune 100. It used "AI" to analyze employee relationships and determine which employees would form the best teams etc...
Other day my boss asked for me to ensure I'm accepting or declining meetings. As a PM with 20+ mtgs a week I default attend without accepting as constant rescheduling and re-accepting is not productive. I've done this for 2 yrs without issue... Now it's an issue.
I also write most of my docs in Pages and export to Word. Guess I'm not getting my full potential productivity score for not using MSFT.
Speaking as a newish remote worker, it's a pain in the ass to schedule time with people who don't accept or decline their meetings, because when you go to schedule something with them, it's all an undifferentiated block of tentatives, and then I have to spend even more time than the person saved by not simply clicking accept or decline by bothering them through email or slack.
Also, once it's known you do this, you will find yourself with people double+ booking you because they know your availability information is meaningless.
I don't mean this to sound hostile or presumptuous, but please reconsider whether this policy is foisting off a negative externality on your colleagues and/or actually saving you time.
If you never reject or accept meetings, a colleague can never know if a booked slot in your calendar is something you intend to attend or just something you don't care about but didn't reject. So your availability in your calendar is meaningless.
I have colleagues that permanently have their entire calendar booked, even if they attend only half of it. So either you send them an invitation regardless of their availability, call them to check, or leave them out of the loop entirely.
Why would you invite someone to a meeting if you don’t expect them to attend?
At least that’s how it works in our company, so if you’re invited you are expected to attend (and be unavailable) unless you explicitly reject the meeting (which I don’t think anyone but me ever does).
There are all sorts of “meetings” that are optional- largish seminars/brown bags/trainings, meetings sent to multiple members of a partner team with the expectation that only one representative attend, social events, etc.
I’m not sure you are indicating that people or shouldn’t explicitly accept or decline invites, but I’m definitely in the “keep your calendar accurate” camp (although my orgs cultural expectation is that “tentative = free”)
It defaults to tentative if you’re not accepting the invitations. So yeah they’ll try to double book you because you get to tentative by either being listed as optional on the other invitation or being non-committal.
Also the organizer of the original invitation never heard back from you so they don’t know wtf ur doing.
Accepting/rejecting may not be valuable to you, but it’s a courtesy and a time-saving tool for organizers.
You know your own workflow better than any of us.. but I'd also be willing to wager that you're losing minutes responding to inquiries about your true availability in return for not losing seconds clicking "accept" or "decline" on meeting invites.
I default attend without accepting as constant rescheduling and re-accepting is not productive
I started doing this for the same reason but it resulted in another problem: meetings started getting double booked or even booked in triplicate.
So now I actually “Respond without comment” so the person at least knows I’m attending and presumably other people will know I’m in a meeting already at 1pm.
Yet people STILL seem to adamantly refuse to actually look at the scheduling assistant in Outlook.
One side effect of covid is that I’ve developed a very good skill of enforcing my availability to others who seem just fine trampling over the best laid plans without even TRYING to communicate about it.
Or when they do, it’s:
“Hey your calendar says you’re out on Friday, do you have 30m for a quick zoom Friday at 2?”
This happens to me quite a bit and I used to wonder what was going through people’s minds. If my calendar shows I’m busy or out of office, then I can’t meet.
But then I asked some people and they told me that most people’s calendars are not dependable and have all sorts of garbage. People have “auto accept” turned on and all sorts of weird behaviors.
I try to keep my calendar accurate and either accept or decline real stuff and ignore/leave tentative things I don’t care about.
My normal line is “schedule any time free on my calendar that’s convenient for you, ignore any tentative.”
To be fair, when I was an IC it was far easier to decide the day of if I was going to attend a meeting because aside from standups and sprint planning/retro, meetings were few and far between.
Nothing against that, but that’s when tentative makes more sense. I tentatively accept many items for similar reasons, but I work to always attend stuff I accept.
> I've done this for 2 yrs without issue... Now it's an issue.
Without an issue to you. Has it make work harder for your colleagues? I try to not assume that just because I can’t detect an issue doesn’t mean that they don’t exist.
I guess your contributors should stop telling you whether they're ok with your deadlines or not also, and leave the surprise of whether it's done or not for the moment you actually expected it.
Personally, I think changes like this are good and necessary to educate users on exactly what's happening to their data.
Eventually we'll hit a critical mass; I am already thinking about how to (I love snails) confuse AI, pollute it with bad data and make it absolutely useless.
Maybe I'll send emails (hello to the AI!) that make no sense, tag random people and make the analytics as useless as possible.
We should all be uploading a few TB of random junk to Google photos, sending 1000's of pointless emails to random people and storing as much useless data as physically possible that they won't know whether to keep or purge.
At least if they're paying to store my 1PB of absolutely pointless data that they can't do anything with it'll make them think about whether it's worth the cost.
I worked in a kind of work-hard-work-hard place (finance). People often making a point of how laate they worked etc.
The most productive software guy, by any reasonable metric of quality and speed, was a guy who came in at 8.30 and left at 5 Pm on the dot every day. He avoided meetings if only could, replied to emails in batch at the end of the day etc. He would suck at these metrics, and yet he’s hands down the best SE I ever worked with.
For many people staying late is a coping mechanism - they want to escape their abusive partner or mounting chores waiting to suck last breath out of their miserable lives. How good employer could tackle it? At one place I worked, there was a hard 5pm cut off and regardless what was happening people had to leave the office. Any temptation to stay late was nipped in the bud. I can imagine people would still continue working until late from home anyway. If work is an escape, people will find a way. I learned one employee would take their laptop and go to a nearby pub and work in the secluded area of the premises. He said he could do either that or end it once and for all. Didn't want to hear about therapy.
It's really sad. We already have an incentive to game other people's perceptions to get an easy promotion but this takes it to a whole new level, and seems to reward the trickster political over the people trying to just make an impact.
My expectation is that it's going to damage the productivity of the most productive but slightly increase productivity of the laziest quartile who should've be fired anyway. The biggest impact will be to drive engagement with Microsoft products which was the real endgame.
"What's Work Insights?
Work Insights is a reporting tool for G Suite that gives you insights into the impact of your G Suite deployment. Using easy-to-read charts, you can see metrics on your organization's G Suite adoption, productivity, and collaboration."
So they are putting "productivity" there between "adoption" and "collaboration". This tells me that it is selling usage metrics which might be the same thing, but I don't see an aggregate score per employee named "productivity xxx" which is the big issue with MS product.
Also google's metrics are anonymous (see "Work Insights and user trust"). This seems to be a big difference.
Looking at metrics is fine. You need to know how the tools you are paying for are used. Grading employees according to those metrics is bad.
If my job doesn't involve using any Microsoft Office apps other than teams for communication then my productivity score should be zero? I should appear as sitting around all day and occasionally having meetings
I talked with an internal dev group that used all Visual Studio dev stack. They thought that everyone on the 25k org should code their way. It was weird, I thought they were joking, but they were serious. They said with a straight face that no one should ever need to write in Ruby, JavaScript, or Python. When I pointed out that most major ML packages were in Python (they were into “AI” and whatnot) they said that PowerBI and Azure could do that for them.
I kind of wish I was so certain about things in life. This was also a few years before Microsoft bought GitHub and they talked about how no one needed to use Git and that Visual Studio had built in version control that supported git, but no one liked it.
I kind of wonder what they’re up to now as Microsoft is shifting over to GitHub. That was part of my worry with the acquisition is whatever cult-like mindset from Microsoft will bleed over.
Sounds like a good way to pick up "loyal" users. If they weren't using your software suite before (eg. through educational licensing programs), they will have to switch to it now, if they want to look "productive".
It's a totally useless metric for productivity but I think the adoption thing is something that should be tracked and used. I big corporates you have a lot of holdouts that will send Emails instead of using a shared Doc for the next 10 years if you don't somehow tell them to not do that.
And no question this has nothing to with productivity or an individuals impact on the company, but I think we here seriously underestimate ho slow some people are in using new available tools and I can see some value in that perspective.
AFAIK, no. Also, as a programmer, a lot of this stuff doesn't necessarily correlate very well with my actual productivity.
My personal views on this are that this is a pretty invasive feature and I don't think that this feature is ethical. Maybe if it was really clear about what it was doing it would be ok, but even then I still don't like it.
Some stupid MBA with a bow-tie was probably pitching how this is going to be a "game changer" for the product. This is already thanksgiving, so probably he collected his nice bonus for "innovative contributions to the company" in reality it's just a disgusting weapon for management to bring it up as part of your 1:1s with your bosses or when you ask them for a raise.
This sounds stupid, but some of this stuff seems useful to focus trainings.
One of my pet peeves is when someone takes a onedrive doc, downloads it, edits it, and sends the off cloud copy back by email. Office sucks at merging documents so this is more work for all the other “cloudies” or whatever you call people who know how to edit online.
I’ve found this is because some users just never used cloud document tools and don’t know how and there’s no easy culturural norm now that everyone is teleworking to teach them.
But this should be used to find zero or very low users and then thrown away. I think that trying to see the number or make a metric out of “number of opens” is so phenomenally dumb. It would be like comparing two git commit history grids and thinking one person is better than another.
I am actually building something even worse but in parallel I am thinking that the same skillset used to build monitoring could be used to build a tool to game these tools. Comment here if you are interested in building the roboemployee of the future to be used by employees rather than employers.
Bloated code metastasizes bugs in badly architected codebases. If a project like this survives long enough, some developer will suffer the consequences. I had to deal with one of these projects recently and it's making me want to switch jobs.
Considering I spend almost all my time between various development tools and almost exclusively use wiki for documentation, I'll probably fare terribly by these metrics. Few days per year spent in Visio being the only saving grace.
Most likely, if companies start considering this as a real metric for raises or promotions, someone will come up with ways to fake data to increase stats. Besides, it doesn't seem prudent to gamify something so relative.
I hope that Microsoft knows that using software or cameras exclusive for employee surveillance is illegal in Switzerland, but if they sell it, no one will do anything. Just that one data protection officer will get angry.
Just before I left my last job a month or so ago, I checked my Github stats. Over the course of 8 years I added 124,000 LOC and removed 466,000 LOC. I can also assure you that my employer was sad to see me go.
Have these metrics only just been rolled out to some places? I noticed I started to get them at my work in the UK about a year ago. I did think they were pretty weird. But why are they suddenly being talked about now?
I predict that Microsoft will rename that "Productivity Score" to something boring like "Product Feature Granular Usage Report" and do nothing else, and press will stop covering it.
The topic of work productivity is interesting from a behaviour science point of view. One wonders how far we've come in what is effectively manipulating humans to work faster.
You get what you measure, so, good luck with that.. When they've made the system perfect, the system can do the work all by itself, that's probably an even better biz.
Can I break topic and complain how absolutey goddamn awful this is to read on mobile? Literally 100% of the screen before I’ve even seen the headline is an embedded ad banner, stacked on top of a video ad widget, with a hovering ad banner down below.
Scroll down, headline, one sentence, nearly full phone display length ad. Scroll more, one paragraph another Nearly full length ad.
I saw that recently(AdGuard). I use both ABP and UBO for desktop. Often I'll promptly back out of a website due to the aggressive ads. Funny living without ads so long get triggered seeing them ha.
If you can drill down to identifiable users. Then I have a hard time seeing how this is not a blatant GDPR violation. The biggest GDPR fine so far has been over workplace monitoring. So brave of Microsoft to roll out something like this, this might end up being a costly move on their part.
It's been 3 years now and still when I see Gizmodo I think "the folks who published Damore's report while removing all references from it and without informing the public they did so".
Maybe I’m weird but none of this really bothers me. If a company is shitty and micromanages and bases promotions off this crap then they’re a shitty company. And they’d be a shitty company whether tools give them extra data or not.
On the other hand it might be good if a company doesn’t suck and they use use this data to assess whether remote work is causing a decrease a work/life balance or causing people to spend too much time in too many meetings.
Data is just data. If a company is using data for nefarious purposes then they’re probably just a bad company to work for.
I can go to places that don’t use this type of data to make bad decisions.
Some places give employee ratings based on number of bugs fixed, lines of code written, or commits made. This is stupid. This data is also available to basically all companies. The fix is NOT to block that commit or bug counts from being visible. The fix is to not over index on that data when making rating or promotion decisions.
Just like you can go to companies that don't have open office plans? Sometimes these pernicious fashion trends spread so far and wide that in practice you're stuck with it
Open office plans benefit the employer. It saves them money. Using bad data to make bad decisions about employee careers does not benefit the employer.
Most of this information is already available. I don’t anticipate this actually being a problem. I could be wrong! But so far in my 14 year career I’ve not been held back by bad decisions based on bugs closed, lines of code written, or diffs landed. I do not anticipate a new set of even less relevant data changing this.
> Using bad data to make bad decisions about employee careers does not benefit the employer.
It's another useless number that they can pretend to care about when negotiating salaries in order to keep down wages.
- "Weeeellll... I can se that your Productiiiivity score is just at 74.3%, the average in your team is 77.6% so I'm affraid that it's really impossible for me to increase your salary as things are looking out. I'm reaaaally sorry. Have a good one!"
Agree with your first paragraph. Regarding whether it'll be used in a deleterious way I guess that'll depend on the boss. I could see it making bad bosses even worse but leaving good bosses unaffected.
> I can go to places that don’t use this type of data to make bad decisions.
Now re-read the post you replied to.
> Once everyone starts doing this there will be no where else for you to go.
No, you will not go anywhere where this practice is NOT done, because it will become widespread and you will have to be really lucky to find such a place where those practices are not done. That is, if they set a precedent and companies follow and people tolerate.
> probably just a bad company to work for.
Shit will happen when the majority of the companies will be bad companies to work for.
Lots of irrelevant data is already available. It has never been used against me. The presence of additional irrelevant data does not seem likely to change this.
Could be wrong! Feel free to ping me in 5 years and we’ll see if I regret everything. But I don’t anticipate significant grief over this.
I have worked for a company that builds a tool like this. The company is founded by ex-Crossover (look up the company called Crossover and what the employees say about productivity measuring over there) executives with a goal of coming up a better productivity measuring tool than what they've already had at Crossover.
With their own tool, they were measuring everything from my keystrokes to random web cam/screenshots in every ten minute interval. They've given nice names for them like "Focus" and "Intensity". "Focus" is a measure of how often I'm switching context and "intensity" is a measure of my keystrokes/mouse clicks/mouse scrolls per minute.
The surprising thing is that they want to market this product to knowledge worker environments and there were already about 20 competitors for them, which says this has a future. In fact, every software engineer/marketeer/PM in this company were already being evaluated partly - mainly, when they want to bash you in a 1-1 - based on these metrics.
You're right. I only worked for this company because I needed to make a few bucks. I would never go back to work somewhere if they use a tool like this.
>> In IBM there's a religion in software that says you have to count K-LOCs, and a K-LOC is a thousand line of code. How big a project is it? Oh, it's sort of a 10K-LOC project. This is a 20K-LOCer. And this is 5OK-LOCs. And IBM wanted to sort of make it the religion about how we got paid. How much money we made off OS 2, how much they did. How many K-LOCs did you do? And we kept trying to convince them - hey, if we have - a developer's got a good idea and he can get something done in 4K-LOCs instead of 20K-LOCs, should we make less money? Because he's made something smaller and faster, less KLOC. K-LOCs, K-LOCs, that's the methodology. Ugh anyway, that always makes my back just crinkle up at the thought of the whole thing.
[1] https://www.pbs.org/nerds/part2.html
[2] https://www.youtube.com/watch?v=kHI7RTKhlz0