Microsoft Productivity Score Feature Criticised as Workplace Surveillance (theguardian.com) 60
Microsoft has been criticised for enabling "workplace surveillance" after privacy campaigners warned that the company's "productivity score" feature allows managers to use Microsoft 365 to track their employees' activity at an individual level. From a report: The tools, first released in 2019, are designed to "provide you visibility into how your organisation works," according to a Microsoft blogpost, and aggregate information about everything from email use to network connectivity into a headline percentage for office productivity. But by default, reports also let managers drill down into data on individual employees, to find those who participate less in group chat conversations, send fewer emails, or fail to collaborate in shared documents. "This is so problematic at many levels," tweeted the Austrian researcher Wolfie Christl, who raised alarm about the feature. "Employers are increasingly exploiting metadata logged by software and devices for performance analytics and algorithmic control," Christl added. "MS is providing the tools for it. Practices we know from software development (and factories and call centres) are expanded to all white-collar work."
It's more like a joke (Score:5, Interesting)
I get feedback emails from Microsoft productivity at work, and it's not even close to what happens. As far as I can figure out it takes the percentage of time that you work on emails and teams and calls that productivity and then send reports which reduce my productivity because I have to read them.
Re: It's more like a joke (Score:3)
And this is the predictable outcome to something like this.
Oh well, this just makes people feel less bad about helping themselves to the contents of the supply closet.
Re: (Score:3)
Indeed. It is another worthless and counter-productive counting-metric, obviously thought up by complete morons.
Incidentally, does anybody know how to stop these? I currently get 3 (!) different ones on three different o365 accounts.
Re: (Score:3)
Re: (Score:1)
It's worse than a joke, it's marketing (Score:5, Insightful)
Microsoft doesn't care about your productivity. They want to increase your usage of Microsoft products, and this metric will do that. This even feeds into PHBs' beliefs that using these tools is a sign of a productive employee, and they'll put a goal on your annual review to increase your score. It's a brilliant sales tactic.
Re: (Score:1)
Re: (Score:2)
I get feedback emails from Microsoft productivity at work, and it's not even close to what happens. As far as I can figure out it takes the percentage of time that you work on emails and teams and calls that productivity and then send reports which reduce my productivity because I have to read them.
If I have a routine task to do, and I automate it away, for example, by provisioning about a million servers using scripts, and my time working on the scripts is about a day, will it show as a productivity fail because I only worked a day, even though I did the boring task a million times, automatically? Analytics may not account correctly in this regard. Software Development and System Administration are the most probable to give these kind of false-positives.
Re: (Score:1)
Re: (Score:2)
You seem to be confusing productivity at work and productivity FOR MICROSOFT.
Your time using Microsoft products is immensely productive for Microsoft. The time you are not using their products is not at all productive for Microsoft.
Participate ... (Score:3)
... less in group chat conversations, send fewer emails, or fail to collaborate in shared documents.
Sorry I missed the office football pool. I was working.
Finally... (Score:5, Insightful)
Ask stupid questions, get stupid answers (Score:1)
Re:Ask stupid questions, get stupid answers (Score:5, Interesting)
After about 3 weeks I just got tired of it and marked it complete
One really has to wonder what is going on in that organization. Marking something complete means that the process downstream of your task that expects it as an input is going to fall flat on its face. A couple of anecdotes:
I was working in a software group where we had to write test routines for automated test equipment. One guy made schedule by cobbling together a couple of subroutine stubs that just returned a 'passed' status. And that's what made it into the production code. The first time his modules passed a defective unit (which was discovered downstream), and his work was revealed, he was marched to the gate by security.
We also had a system called Engineering Scheduled Work Release (ESWR) that tracked and reported dates of drawing submittals. One group, falling far behind, just released a set of drawings, all marked "This page intentionally left blank" just to make their release schedule. Needles to say, the shop would have come looking for their heads if they were handed such input. Fortunately, the program was so f*cked up, they easily managed to submit drawing revisions with an actual system design long before it was actually needed.
Re: (Score:1)
One group, falling far behind, just released a set of drawings, all marked "This page intentionally left blank" ...
Was this Trump's post-election legal team? :-)
Re: (Score:1)
I've seen variations of this theme in quite a few different places, some of them even blue chips in the IT sector. As soon as MBAs want shiny colourful dashboard to make decisions that no one needs, the engineers will come up with solutions like this. I squarely blame the MBAs in middle management here.
Re: (Score:2)
One group, falling far behind, just released a set of drawings, all marked "This page intentionally left blank" just to make their release schedule. Needles to say, the shop would have come looking for their heads if they were handed such input. Fortunately, the program was so f*cked up, they easily managed to submit drawing revisions with an actual system design long before it was actually needed.
This is the Schedule Chicken anit-pattern. Quite common in huge projects with crazy deadlines. Instead of taking the blame for delaying the schedule, pretend you are on-track and let other teams take the blame, the use the extra time to actually finish your task.
The more messed up the project was, the moer likely your downstream tasks would not be able to start on time.
Re: (Score:2)
One really has to wonder what is going on in that organization.
Apparently it was the sort of place where management would rather continue demanding the impossible rather than admit they may have goofed when they made the timeline.
Re: (Score:2)
Excellent. Always nice to see there are smart people around.
Re: (Score:2)
This happened to me as well. A particular manager wrote his own scheduling software and insisted we use it. It was awful beyond belief. We literally had to e-mail him to request any changes to be made, and then a day or two later it would show up. It was so tedious to use that we eventually just stopped requesting any changes, and left the schedule completely alone. Naturally, after only a month or two, it bore no resemblance to reality. When a task came due, we just dutifully marked it completed.
Natu
Re: (Score:2)
Keeping two sets of books [cfainstitute.org]? The circle is now complete. Or maybe it's more "cyclical".
Re: (Score:2)
I used to work at Nortel doing source code control for a number of various projects. One day I get an email from some manager asking why I hadn't implemented his request to put a graph of the historical number of lines of code into the weekly build report and the message was CC'd to just about every VP in the company.
As I'm trying to figure out how to diplomatically tell this idiot to f-off my manager, who was copied in the message, walks over to my desk, quietly says that he'll handle it, and I never hear
Anonymizing the details (Score:4, Informative)
In the Settings area of the Admin Center go under Reports and select “Display anonymous identifiers instead of user, group, or site names in all reports”. This should solve the problem if your organization finds these details too sensitive and does not wish to collect them without anonymizing them.
So just ruin the data such that it is useless (Score:5, Insightful)
Seems to me to be very similar to how GitHub reports 'productivity' via how many commits you make, which is a laughable way to measure actual useful work as it completely excludes the mental-only side of good software development.
So someone made a quick/dirty tool to make artificial 'work' happen so you can make GitHub think you're a monster that does 10,000 commits a day if you wanted. Go look up 'greenhat' on GitHub.
I doubt it would really take much effort to make a similar tool for O365 if someone bothered to look into it.
Re: So just ruin the data such that it is useless (Score:5, Insightful)
Re: So just ruin the data such that it is useless (Score:4, Interesting)
Or rather, it can't be measured, so you're using a proxy which can. The problem is, if it can be measured, it can be gamed.
It's why SMART goals only work for things that are directly measurable it will fail if you try to use a proxy instead as the proxy will measure you getting towards the end goal when in reality it's just people gaming the proxy.
Re: So just ruin the data such that it is useless (Score:2)
"Just"
And when do I work, in-between concerning myself with stupid games like this all day?
Re: (Score:2)
The GitHub graph is useful to know which days of the week people are active and to set your expectations for how long somebody might be likely to reply to you.
My email isn’t in Outlook (Score:3)
And Office doesn’t really have anything to do with web development or Linux system administration. So if our department head brings this up, I’m going to laugh in her face.
Re: (Score:2)
Re: (Score:2)
I'm quite sure that in some areas "knows how to do the work herself" is quite useful. In others, it's totally unrealistic and irrelevant. One team I was on had curriculum designers, a technical curriculum writer, two graphic designers, one of which was a UI designer, a programmer and a half, a sysadmin, and a project manager. The job of the team's boss was mostly to interface with his counterparts in other parts of the agency and made sure we had the breathing room we need. He did a great job. How the he
* a malware sample (Score:2)
Missing word due to a typo:
I don't need my boss to try to tell me how to analyze a malware sample. - I need my boss to deal with the accounting department on their policy for how they (mis)handle sensitive employee data, and back me up when I tell the CIO something he doesn't want to hear.
I need my manager to be good at managing. I'll handle the decompiling and the database k-anonimity, thank you.
Re: (Score:3)
Re: (Score:3)
I think we agree the boss who stays out of the way allows you to get the work done.
In my experience a boss who doesn't know my job, and KNOWS that they don't know my job, has tended to leave me alone and let me work.
The worst are people who think they know, but don't actually know anything. Or are trying to prove that they know, when they clearly don't.
Either way, I think a micromanager is a problem. I say that as a former micromanager myself. I literally lost millions of dollars by failing to delegate appr
Re: (Score:2)
Re: (Score:2)
That depends on what you mean by manager.
If the manager and the workplace sees their job as facilitation and communication, they only need those skills. They will need to defer to the people who actually can do the work for things like pre-requisites and timeline.
If the manager or workplace sees the manager as the boss that gives the marching orders, they need to understand what they are ordering people to do if their deadlines are to make any sense at all.
Note that for the former, manager is a lateral move
Re: (Score:2)
Agreed.
I'd expand on that to say that separating the two roles probably works better in many cases. Actually there are three distinct types of tasks. Frequently one person isn't great at all three areas.
One job is interfacing with other departments and up the org chart. Getting the needed budget, based on presenting the value of the team in business terms. Dealing with unreasonable demands from other departments. They make sure the team has what they need to get the work done..This is a people job, polit
Re: (Score:2)
Soviet Shoe Factory Principle (Score:2)
https://www.openpolitics.com/l... [openpolitics.com]
Re: (Score:2)
I heard a different take on the anecdotal Soviet Factory vs Metrics.
A factory that produced nails was told they needed to ship a specific quantity of nails every month. Result: the factory started manufacturing thumbtacks. When the higher-ups realized their mistake, they switched the metric to measuring the *weight* of their manufactured goods per month. Result: the factory switched to manufacturing railroad spikes.
The point is the same, of course: If you introduce artificial metrics, people will find w
Re: (Score:1)
They are not "artificial" metrics, they just shouldn't be over-emphasized. For example, lines-of-code is a useful metric, but if you mostly reward based on it, coders will produce bloated verbose code. Same for number of screens or number of entities worked on.
Good evaluations of non-trivial work probably require monitoring at least a dozen metrics, many of them either subjective, or too costly to clearly quantify such that the org spends more money measuring then doing.
[Corrections] Re:Soviet Shoe Factory Principle (Score:1)
Re: "spends more money measuring then doing."
Should be: "...than doing."
Re: "Same for number of screens"
For example, coders may be less likely to look for ways to consolidate screens if they are rewarded for screen quantity. This assumes the coders do hybrid programmer/analyst work. In some places, coders are not allowed to make suggestions, which is dumb, but happens.
Re: (Score:2)
Frankly, I'm not sure that measurable metrics, no matter how many you choose to evaluate, can ever really encapsulate how well or poorly a programmer is doing their job.
I think they can certainly be used as indicators, but the problem with metrics is that they have a way of missing intangibles. Say you have a programmer who is an excellent mentor, and helps younger programmers develop good coding and problem-solving skills, thereby increasing productivity for the entire team. How is that measures? Or, sa
Useless irrlevant metrics signifying nothing (Score:5, Interesting)
I worked for Intel for several years. Lab management people wanted to better allocate expensive equipment like oscilloscopes, so they installed software and USB WiFi adapters in them all to track their 'usage'. How they tracked the 'usage' of the 'scopes was by tracking keypresses; in their judgement if a 'scope sat there for days without having anyone push any of it's buttons, then it clearly must be 'idle', therefore can be scooped up and trundled off to someone else for them to use. The people managing lab equipment, being paper-pushers and bean-counters, and therefore having no clue whatsoever about what the 'scopes are even used for, and more importantly, how they're used, had no clue that front-panel buttons are rarely used since all the testing they're used for is run by automation software. Few if any button-presses, and there were in fact stretches of time where no testing was being done, but that didn't mean the engineers and techs using them had no further use for them. Wrong metrics being tracked by clueless people who had no idea what 'getting the work done' actually meant.
That's what most 'productivity tracking' like in TFA is like: worthless, pointless, unrealistic, and only appeals to obsessive-compulsive anal-retentive paper-pushing bean-counting management types who want to be looking over people's shoulders all day every day, trusting their employees about as much as you trust convicts in a maximum security prison.
Here's a suggestion to management types: How about you judge the 'productivity' of your employees by how much they get done and whether it's done in a timely manner, and not based on 'busywork' nonsense? If something doesn't get done, then ask them why; there's likely to be a reasonable explanation, since shit happens, and as Murphys' Law of Combat states, 'No plan survives contact with the enemy'. If someone consistently underperforms, then you can be up their ass all day every day, they'll either shape up or you ship them out, but annoying the actually productive, hard-working, responsible people benefits no one.
libreoffice (Score:2)
Last place I worked, "The Dungeon" used Office365, but I did all my work in Libreoffice and emails in Thunderbird and avoided teams, preferring to go and speak with people (what a yesterday concept).
ROFLMFAO... (Score:5, Interesting)
Group chats, emails sent, and shared docs dithered in? That has got to be just about the most stupid and worthless measures of "productivity" I've ever seen... even dumber than lines of code. I guess it's par for the course though, for the sort of PHBs that would want to use Microsoft 365 or any other sort of workplace "productivity monitoring" software in the first place. I routinely work full days without wasting time on Slack, a full week without sending emails which will never be read anyway, and a month without nitpicking at someone else's grammar and spelling mistakes in the documentation.
Just give me my OKRs and deadlines and then GTFO. The work will get done.
Re: (Score:2)
We get those reports, and our company is standardized on another non-MS chat system, and uses non-MS meeting sites 2/3 the time. So it's a metric of what remaining is done using MS products... for most employees, others like me are paid to work at Linux and Unix command prompts. We get very low scores, haha
Measuring Noise as Datum (Score:1)
Wow, for a company with "smart people" the measuring noise in the veneer of gathering data is incredibly muddle headed. And their data analysis, tools are only as good as the datum collected...but the noise is...doesn't surprise me. This approach has the veneer of management, but is micromanagement and control. Signs of a toxic culture underneath the guise of egalitarian policies...a Potemkin village.
This reminds me of in Halberstam's "The Reckoning," how in one auto company they had a glass toilet, from t
See Goodhart's Law... (Score:3, Insightful)
"When a measure becomes a target, it ceases to be a good measure."
Eventually workers will clue in, and start producing low-effort work to increase metrics, ultimately lowering SNR and decreasing real productivity.
Same old dance (Score:2)
Workplace surveillance is as old as work itself. The only thing that changes is the technology,
The four laws of workplace time management
1) Management love it
2) Employees hate it.
3) At least one way to successfully game any surveillance system is discovered within 4 hours of implimentation.
a) It will spread throughout the department within 2 weeks.
b) Moving forward, new and better ways will be found every two weeks.
4) Given the choice,
Ironic considering... (Score:2)
Pretty ironic considering it's usually Word that kills my productivity when it screws up the formatting of my document and the useless "show codes" does nothing to help solve the problem.