The Broken Microsoft Pact: Layoffs and Performance Management
danielsada.tech
danielsada.tech
I think humans are fundamentally flawed in not being able to see alternate history. If they have to pay a union, they will not see all the benefits, and only focus on the 50 dollar union fee.
This was a time where companies were hiring talented developers just to deny them to their competitors. Nobody was interested in shaking up that status quo by becoming part of a union.
Just look at European government run pension systems which are basically legal pyramid schemes, that everyone knows are unsustainable the way things are going right now, but nobody wants to do anything about it when it's more convenient to kick the can down the road until it become a problem under someone else's watch.
also keep in mind that unions first popped up in places were people were, like, dying, or getting maimed: trains, mines, electrical, gasfitters, ironworkers. up until now programming has been easy money sitting indoors while sipping lattes.
The first unions aren't the only kind. Unions benefit all sorts of fields. They continue to fight on behalf of labor. In all fields, management and labor are at odds and management has a collective bargaining voice so labor needs one too.
Sorry, you misspelled "decades of Republicans disassembling labor protections and enacting garbage like right to work laws, corporations hiring thugs like the Pinkertons to break up and discourage them, and conservative media spewing lies to discourage membership".
Both parties are so in capital's pocket that, at least on labor issues, they are essentially the same.
Yes, Biden could have done better on railroads, but his administration:
* Made millions more workers eligible for overtime pay
* Installed union leaders on the NLRB (contrast to Republicans who fired members, defunded efforts, and gave access to NRLB data to Musk/Palantir/etc.)
* Those new NLRB members issued the Cemex decision making it harder for companies to suppress union voting
* Issued any number of rules in favor of workers i.e. https://www.dol.gov/newsroom/releases/olms/olms20230727
* Raised the contractor minimum wage
* Increased funding for the NLRB
* Ensured that companies couldn't avoid penalties by forming smaller companies - a process the Trump NLRB created to make it harder to penalize companies unionbusting
This is only a few things. There is a world of difference.
And this is more true now than ever before since they can (so far legally, if not morally) also use LLMs to whitewash off GPL and other such licenses that would in the past have put practical limitations on their usage.
If you've built up expertise in an open-source projects that lots of other companies use that knowledge is a valuable asset on your resume.
OSS? You mean copyleft? Who does that? Who would use that?
IIRC, only in France unions can prevent companies from doing redundancies if their bottom line looks good, but that's probably also why many companies aren't hiring much in France to begin with, why skilled wages are lower and youth unemployment higher there compared to other equally developed economies like Netherlands or Germany for example.
I enjoyed many years the agreements of IG Metal in the telecommunications and life sciences sectors, as software engineer.
I got my dues’ worth in just the agreement they quickly got to during the post-COVID inflation to adjust our wages. Of course, I would have still gotten it had I not been a member, but it takes people willing to be members to make them (us!) strong enough that employers take them (us!) very seriously.
Oh, and the pay is way better than a lot of IT consultancies here, not to mention the hours and general working conditions.
Most of the issue is that there's too much administrative policy (whether the imposition is internal or external doesn't matter) for us to effectively communicate and collaborate. Unions would only add to that while collecting fees from us. Most of us are intelligent enough to know this which is why we never form them.
Question for you about this. Is the balance of power in negotiation equal between a lone developer and a company?
I ask, because in my little world, it sure seems like the company has way more of the cards than a developer does.
Unions could only make that worse.
But is it equal? It is hard to scientifically to prove, but it sure seems like companies have way more leverage than a lone developer.
I am not trying to antagonize you, btw, I am just seeing where you come from in your points.
Or like teachers in new York getting paid to sit in the rubber room as union fights for months to defend them after chronic drinking at work. I miss my beers in my terrible non unionized workplace.
And it's not like I plan on doing anything illegal, but if I ever misuse user data or whatever it would be nice to have a wall of silence (just need a color, blue is taken), instead of all those pesky whistleblowers.
I would especially love to be paid based on seniority and have (the small minority, but still) useless, lazy and incompetent coworkers I had keep their job and be paid the same, especially the guy I repeatedly caught playing fantasy baseball. He got fired (I didn't say anything, his lack of performance was also pretty obvious), I think it would be much more equitable and would really motivate me if he got to keep his job and was paid same as me.
Where do I sign up to pay only a small fee for these benefits?
But regardless, American style unions are just a uniquely malignant institution. They do/did unions better in Japan for example or even in Europe; although I still think these are a net negative, compared to them US style unions are just tailor made to suite parasites and criminals. Or as they say, my opinion of them was better before I heard of the union-mafia connections, but now I've lost all respect for mafia.
This is the worst part of averting disaster - people never see you fixed something before it became a problem. That's why there are so many "disaster-driven organizations" out there, where the people who prevents disasters gets passed over by the people who often provoke them in order to be the hero that fixes the problem.
Given what's been happening in the tech industry and the economy at large, I now keep a 1-year emergency fund in a money-market fund at 4% (Fidelity SPAXX). I'm probably losing out on some growth (SP500 grew 11% over the past year, despite the massive drop in April 2025), but at least I have liquidity in case I get laid off.
That's the kind of game I feel I have to play these days.
https://fundresearch.fidelity.com/mutual-funds/summary/31617...
But if you have much more than that, there’s no reason to keep an entire year in such a low return investment.
Money in say an S&P500 ETF can be liquid in 1-3 days.
We went through this in 2022 when most stocks were down 20% for a year. SP500 is not a low risk short term investment.
How would you feel if your rainy day fund lost 20% of its value?
It's not a sure win rule.
A better strategy is to hedge (bet maybe 3 months of your 1 year emergency runway), but that requires some acumen.
Yeah that’s what I was suggesting when I said you don’t need keep your entire 1 year emergency fund in low risk investments (assuming you have much more than a year of runway).
Nor is it something unique to them. As far as I know, the only large US tech company that didn't do layoffs in the past decade is NVIDIA (their last one was in 2008).
That wasn't a traditional layoff - it was a reimagining of the development process and the elimination of SDET which was overwhelmingly a good thing - I also joined in 2009, and SDET was an utter disaster. All the good SDETs got out of that job - either to SDE at Microsoft or to SDE at another company. Those that were left were largely a waste of money, and the entire culture of "this person writes the code, this person writes the tests" meant that a lot of devs got high recognition and rewards for writing untestable unmaintainable garbage that someone else had to try to cover.
Whenever it comes up among my co-workers as a Microsoft product falls on its face yet again, most recently MS Project Online screwing up something as simple as completion percentages during a meeting, I just sigh and quip "Maybe Microsoft ought to consider hiring a QA department."
What I learned is: you need to hold a high bar, because people can do anything to keep their job, and often not what you want them to do
What you want is someone who is open to feedback, understands it, and takes effective action.
Outside of that, there’s a whole gamut of people. Some get defensive. Some are politely open to feedback but don’t actually try to understand. Some understand but don’t care enough to follow thru. And some try hard but aren’t effective. All of that is bad for your team, and unlikely to change. Just need to cut your losses to open your seat for someone who can do it.
The current person i have is open to feedback, but doesn’t fully understand it and doesn’t care to. It’s like dragging a horse to water. After doing that for six months my manager pointed that out, it’s just not a good fit. I like to see the best in people, and even a little bit of improvement gives me hope. But it’s dragging down our team potential. It’s a hard truth.
But hey I am judging off of basically one sentence you wrote so what do I really know about your situation?
It's a two-way street.
And what I’ve found is your team knows who is and who is not performing. And if you fail to do something about the low performer under the guise of being a nice person (or hoping they’ll eventually figure it out), your team will lose respect for you.
A team of high performers does not want to have to carry along a straggler, no matter how nice they are.
In my experience, I wish I’d made those hard decisions sooner in hindsight, rather than hoping they’d get where I wanted them to be.
It can create some weird unintended consequences though. Like if people know you regularly manage out low performers, they might be risk averse to try something difficult for fear of failing and losing their job. They need to be able to see exactly why someone didn’t make the cut.
It’s a difficult line, especially when complicated by blunt corporate incentives like stack ranking and PIP’ing the bottom 10%, etc where it can be less than clear sometimes.
Also Apple is good at lot of outsourced work for long time compared to many other companies. So they don't tons of non-core software stuff in-house. In that matter they are more like typical Fortune 500 companies than SV company.
Amazon is in the same ballpark AFAIK, but the other two have noticeably better comp. The "Microsoft deal", such as it is, was that it's a place where you don't have to grind hard. This has changed now though.
The common wisdom at the time was that different divisions of Microsoft had widely varying cultures, with as much difference as you'd expect to see between unrelated companies, so perhaps that was just DevDiv.
I left in 2018. I was 63 (first level senior) and had offers from Google and Meta for a downlevel (L4) at a 40% increase in total comp. The gap has closed slightly in the last 7 years but not much at those levels.
Some of the companies I've worked the big "family" culture and "our people are our greatest asset" company principles were the ones to do the deepest and swiftest cuts. Meanwhile their supposedly comparatively ruthless shark infested competitors keep on keeping on.
The people I saw most impacted by this were the ones that took the culture both literally and seriously, staying in the big happy family companies long enough to develop far too much company-specific rather than industry specific expertise, just in time to get laid off at 50.
I think this is a great point. Lately this is number one thing I feel concerned about. Though paradoxically they are coming from different direction as pushing buzzword driven development too fast in process making stable products unstable, calling them out of date. And then getting rid of products along the people who worked on them as dead weight.
I have some passing familiarity with how (California) law firms approach firing, which this article gave me cause to consider:
- they do fire unproductive people aggressively (law firms bill by the hour and attorneys are very expensive to employ, so it’s very obvious and financially meaningful to the business when someone isn’t contributing)
- when they fire someone, they’re very secretive about it. The person stays on the firm’s website for months, and if you call HR, they’ll say that the person still works there (they probably do, in some narrow technical sense). This makes it somewhat easier for the person to get a new job.
Also this a nit but the legal protections aren’t meant to prevent arbitrary termination, which is pretty explicitly legal. They’re meant to prevent discrimination.
This seems more of employee's made up rule rather than Microsoft's. I worked in a company in early 2000s' and old timer's told me similar rule "that here pay is less but little work and lifetime job guarantee". It was of course bullshit made up rule. As economy changed not only did they tighten the screws but also had many layoffs since then.
The powers that be realized an anxiety riddle fearful workforce can get stuff done too just fine as long as they don’t have a better option and you dangle the occasional carrot.
So as long as all corporations move roughly in lockstep you can drastically change conditions without much consequences
It’s a bit like the first big news website implementing a paywall was outrageous and deemed suicide. But if everyone does it…
just another tool to push us to work more for less
It don't work, but you better learn it and use it, because it will replace you and also now you got to delivery 30% more
Aside from that, there's also the matter of fewer people tasked with the same amount of things. The official line on that is that AI makes the remaining employees more productive. The reality is that it's nowhere near good enough for many products, so it just means overworked people making more mistakes and burning out faster.
Every MS FTE whom I knew enough to have that kind of conversation with was stretched thin to the breaking point. Now half their team is gone, and the show must go on apparently. It does not seem viable to me, not even with AI.
Indeed, I think their major trick (in America, anyway) is to make a profitable company unprofitable. Much like the bad forms of private equity from above, they bleed a business dry from below.
essentially imo all 'pacts' between employers and companies is going to change because basically the entire category of economic work we are doing as a society will become not necessary over the next decade due to this
its just that software development is going to come earlier due to its critical nature to the roadmap of model improvement and increasing lab research speed. also due to the fact that RL works quite well for software development, including the more advanced applications like model research
also as a preemptive rebuttal for anyone saying i have no idea how swe/ai research works i am a swe who also does ai research work
> The AI efficiency narrative provides the perfect cover for these layoffs. Companies can frame headcount reduction as “leveraging AI to increase productivity” or “optimizing for the future of work.” It sounds forward-thinking and strategic rather than admitting they failed to manage performance or simply want to cut costs. Whether AI actually replaces the laid-off workers’ productivity is rarely measured or proven, but the narrative sells well to investors and the media.
This so