Xsolla fires 150 employees based on big data analysis of their activity | Hacker News Reader

Xsolla fires 150 employees based on big data analysis of their activity

mcvuk.com

69 pointsjoshbuddy82 comments

Dupe (discussed four days ago): https://news.ycombinator.com/item?id=28068959

I probably look like one of these employees. Honestly I don’t ‘work’ that hard.

But I also have a shit ton of institutional knowledge that cost a good decade+ to acquire. And I got there by working hard.

My value add is to break bottlenecks for my teammates. I dig through hard problems that they find intractable. But I don’t grind through the daily drudgery.

I’m lucky in that my manager understands this and values people based on how they fit with the team and I have something valuable to add.

Broadly speaking my job as a programmer is to turn repetitive activities into processes and code. If I’m still “working hard” a year in you should really wonder about my capabilities.

I probably spend half of my time and way more than half of my energy on removing work we shouldn’t even have to do ourselves, first the random sources (machine bookkeeping beats human bookkeeping almost every time) and then persistent drudge work that steals resources from more important matters.

One job where I was mad about not getting any credit for this, I sat and did math and in six months I had saved us one head count, and I don’t know how much stress (seriously bullshit things). I joked that I could not show up to work and still provide as much value as people who they liked.

> Broadly speaking my job as a programmer is to turn repetitive activities into processes and code. If I’m still “working hard” a year in you should really wonder about my capabilities.

As much as I want to agree with this, what I've found happens if I do a particularly good job is... more work. More projects. I have "bandwidth" available to help out in new areas. When I left my last job my manager and I put together a spreadsheet listing all the projects I headed, and it was something like two pages long in excel. Everyone who saw it had their eyes bulge at the size, and I left because I was being really overworked.

But god damn it, the only reason I want to work hard is so that I don't have to in the future!

I'm seriously considering leaving tech because of this.

You get a better idea of what’s going on when you see beyond the one or two projects that used to occupy your time. That can be interesting in a lot of ways, not the least of which is figuring out that some of the problems you “solved” are just thinking too small and thus picking away at a much bigger issue.

Since I get bored easily, I usually value that sort of opportunity. Usually. If I turn into the cleanup man I start to become sour on the idea of people making messes for others to clean up and then I start looking for an exit.

You're right, in that it's great to be always given new work and interesting projects. But unless the completed work can be handed off to someone else what happens (at least in my case) is that I spend more and more of time making adjustments to past work, either due to requirements changes, competition, feature requests, or bug fixes (which is thankfully pretty rare for me)

These things typically arrive in addition to whatever new project my manager has me on, and the more projects under my belt the less time I get working on what they want me working on, and spend more time firefighting. Towards the end of that job I spent almost my entire day firefighting

I’ve slowly been making my exit from pure tech for this exact reason. 2011 got hired as the “tech guy” for an artist that was winning Grammy’s left and right, breaking world records for sales/ streams/ whatever, with an explicit part of the deal they let me sit in on marketing meetings to learn marketing.

Was able to take “marketing” (really more social science), engineering and design to epidemiology, eventually wrap that together to work in politics, now back on infectious disease (COVID vaccinations). I find “tech skills” just being something in my toolbox to enable me to tackle infinitely more rewarding projects.

I've found the best strategy here is to do rate control on your work on the interface between you and your job. I essentially just try to constantly overperform slightly and any gap that widens between my capability and the demands of the job remains hidden unless it's strategically useful to me.

I won't get paid twice as much if I can do my job in half the time, I just get another half of work. If I use the gap between their demands and my capability carefully I can spend only a portion of it to get raises/promotions and spend the rest on fucking off at lunch and just generally having a chill relaxed time.

Yeah I've adopted that strategy for my current employer. It is exhausting in its own way and difficult to maneuver without being blatantly dishonest.

Seems to me that you & GP both have an unhealthy relationship with your employer. I guess that’s the norm for a lot of us here, I just hate that the generally accepted solution is to hack around it rather than fix (or attempt to fix) that unhealthy relationship.

What's to fix? It seems like the prevailing business culture is to optimize the shit out of everything, and for managers it means getting the most work that you can out of the employees that you have. No matter where I go I am expected to work my ass off.

I would burn all of that down, if I could, because that principle needs to change. But then that's "leaving money on the table", and we just can't have that...

> No matter where I go I am expected to work my ass off. > ... > But then that's "leaving money on the table", and we just can't have that...

Sounds to me like you expect to not have to "work your ass off", but also maximize your pay. I just don't see why you would expect these things to be compatible. There are plenty of opportunities to work 30-hour weeks, if that's what you want. But you're not gonna be making FAANG-level comp -- for good reason: you're giving you employer less, so they give you less in return. If you and I worked at the same company, and you were putting out half the output as me, then I would expect to get twice the comp as you. You're trying to bypass this -- produce half the output as me but still get the same pay. I just don't see why you think that's justified.

If you want to work 30 hours, then take one of the many jobs that accommodate that (hmu if you legitimately can't find one) and accept the pay cut. If you don't want to make that sacrifice, that's on you.

That's my $.02, anyway.

I think you're misunderstanding their tactic here. They aren't working less than their peers, they just also aren't working to their 100% capacity at all times because it burns you out. I've done it myself in the past, and it works, for a time, but it's really only delaying the inevitable. The solution for me was to find a different company that valued what I brought to them, instead of treating me like a functionary in their "process". But, that's a hard thing to do, so I can understand developing a coping mechanism.

Generally speaking, I think your comment and its parent have a poor attitude and not one that anyone should emulate. For me, it's about finding new problems and new challenges. I can't stand facing the same problems over and over again so when I do, I automate/fix those problems and then I find new ones to solve.

You should always be "trying". The idea that you corner some obscure but vital part of your company's infra and then just coast is pretty abhorrent to me. Instead, if you find something difficult, complicated and manual, you should make it easy, simple and automated.

I do believe that firing people based on big data analysis is pretty dumb as some tasks require lots of deliberation and thought without much output initially. But the idea of working hard = being dumb, seems like a pretty creative way of justifying being lazy.

The people that have the bad attitude at my job are all of the burned out people who are now dealing with a list of problems most of which I have been bitching about for years that somehow became a giant avalanche this year.

Everything is fine is a rotten attitude when you are piling up accidental complexity all day and victim blaming anyone who can’t or won’t wade through it. If I were king I’d fire every single person who was happy with the state of affairs a year ago. And then I’d start calling all the people I liked who have left for something better and beg them to come back. We got rid of those assholes, come fix things. Please.

>I think your comment and its parent have a poor attitude and not one that anyone should emulate.

I don't think working hard is dumb. To get into my position I had to work hard and acquire a lot of knowledge / skills. I had to defer a lot of gratification, skip a lot of parties and endure a lot of stress.

But people have lives outside of work. I can accomplish in a day now what might have taken me a month early in my career. I can make better decisions now that result in much less future work... lower amounts of maintenance etc. So should I put in the same grinding effort? That makes no sense to me. I don't grind anymore. Been there, done that.

> I don't think working hard is dumb.

I think GP's attitude (different from 'working hard', more akin to 'overdelivering') can be "dumb", depending on the context. If someone is doing 1/100th of the work per unit of time you're able to do, why would you work the same amount of time for the same compensation? Is that any different than selling gold at the price of copper?

Of course, having equity in the business changes the equation somewhat.

Otherwise, the only issue anyone should have here is what the agreement between you and your employer calls for, and how transparent this relationship is. Just because others in the marketplace are making bad deals shouldn't mean you have to as well.

As someone that enjoys occasionally coasting, I get bored really fast and find new areas to contribute to and problems to solve.

I think you may have gotten downvoted from the tone of the comment by folks that have arcane mastery over some system, which I can get from their perspective, depending on how long they’ve been in the industry/ niche/ company

This may seem horrible but this, I fear, will become the norm. If the boss can’t see you, then they may suspect employees are slacking off at home. So out of caution (or CYA) managers and CEOs will start monitoring all activity available, and transform WFH employees into metrics on a spreadsheet.

If there is a cost savings argument the stock market might even start the drive.

And if it's an effective method of selection, then those companies will thrive, if instead it's a poor method those companies will fill up with busy workers and sink under there own weight.

I look forward to seeing which it is.

Regardless of the outcome it will be spun as a success for the executive team, or if positive spin is impossible they will find a scapegoat or other mechanism to downplay their own responsibility.

Right on... Time spent reflecting, taking a walk to think things through... Running with headphones to keep the blood oxygenated while listening to the latest rambles from higher ups...

You're too optimistic. Companies that are big enough will inevitably do this because it appeals to management, which forms a cohesive and insular culture inside them, and because it can be sold as a cost savings. These companies will inevitably find ways to survive by exploiting their size and power, and so they will be held up as models, and everyone smaller than them will be under pressure to follow their example. The only way it will someday "fail" is if it falls out of favor with management, maybe because some celebrity CEO bashes it in a bestselling book. But after that CEO and his book pass out of the limelight, it will come back, until it falls out of favor again, and so on.

I work at a company that I think is probably big enough, and in a safe enough market that it can't fail. It might not keep up with its peers but competition in entrenched industries makes it hard for a company to fail once they become entrenched. Kind of like barnacles on a ship. The worst that could happen (from an employee perspective) would be being acquired, with the resulting consolidation eliminating redundant jobs.

And many companies are like this. I joke that when our company is trying to make a decision, the first (and often only) thing the executive team asks is "what are our competitors doing?"

Of course it will become the norm. The same people who built today's surveillance dystopia will build this further refinement of it, because "if I don't, someone else will; might as well get paid".

> Jira, Confluence, Gmail, chats, documents, dashboards

Metrics almost as bad as “butts in chairs”, “physically in the office” and my all time favourite “spending 4hrs a day responding to emails”.

I mean it’s possible that these 150 people were doing absolutely nothing… in which case these metrics can be a decent yardstick I suppose, but you can also generate a ton of activity in collaboration tools and contribute very little actual value.

It’s like the help desk tech who just closes the easy tickets because that is the metric they are evaluated on.

As they say, "“When a measure becomes a target, it ceases to be a good measure."

If I worked at a company like this I would spend my time doing visible work that bumps up the metrics regardless of whether that was what really brought value to the company.

-I used to work at a large multinational engineering company which was obsessed by mice - meaningless indicators of company efficiency - for a while, it was decided that we should have large, billboard-like displays at every site showing our progress towards various goals.

We had a (stifled) laugh when our designated goal was to increase delivery precision to 70%[], this was illustrated by a large, easy-to-read poster prominently displayed in the lobby where we received customer representatives.

Sigh.

[] After an outsourcing orgy had made our manufacturing facilities crash and burn, much to the surprise of upper-level management)

This already is the norm. I've been asked multiple times why someone should hire me when they could hire 20 developers from India for the same price. Obviously, that stems from a mindset that assumes that productivity can be measured objectively and scales nicely with headcount.

The correct answer is to wait things out. After they've tried outsourcing for 3 months, they'll be ecstatic that with me, they only have to explain things once.

>If the boss can’t see you, then they may suspect employees are slacking off at home.

If they think employees can't slack off while they're watching, they're probably making too much money for what little management skills they bring to the job.

> Agapitov held a press conference in which he explained that the mass layoffs were caused by the fact that the company has stopped showing 40% growth.

There’s the real story. Everything else is just a polite fiction to rationalize who got laid off.

"This prompted immediate and predictable backlash: both for the layoffs themselves and for the tone of the email. According to ProPerm.ru, the company is investigating to find the employee who leaked the email."

Lol them not expecting such a tone deaf email to get leaked really cements my suspicion that any idiot can find himself in the captain's chair.

I actually respect this email for honesty. Corpspeak gets really tiring after years upon years of hearing it. Here's the corpspeak version of the title: "Sharpening our focus on core principles". Or "Doubling down on X". Would you like to read that instead?

You don't see a third option they could've taken instead?

What are you referring to? Keep in mind this thread is about how to write the email, given the action they took.

How about talking to the employees. How about managers having input on if arbitrary performance metrics were reflective of the value of their contributions? How about not shifting blame on a reduction in growth/ necessity to do layoffs on perf metrics? How about not asking the people you just fired to write you a “long letter” expressing their gratitude?

I agree with all that. This is specifically about the style of writing, not the actions taken.

No it isn't, the two can't be disconnected.

They certainly can. In this case, the message is coming from the top, but the message and the actions are literally disconnected every time a middle manager or team lead has to tell people about some foolishness coming from the top of the company. And when the person making the decision is also the communicator, they can still make meaningful rhetorical decisions such as whether to apologize for their inability to make a good enough decision, to compliment or threaten the existing employees, to attempt to reveal whether to make up for the actions taken in some way, and so on, in addition to less consequential decisions about tone that may affect different employees according to their communication preferences.

I recommend becoming willing to evaluate how bad decisions are communicated to the degree you evaluate good ones as the former exercise aids the latter.

Assume you've decided to do larger scale layoffs. What would you prefer the announcement to look like?

That was my point, why should the layoff be assumed as a given? That was as much of a decision that someone made as the messaging around it.

The article said this layoff started after Xsolla was no longer showing 40% growth.

Having implemented Xsolla, I think they stopped showing 40% growth because the products Xsolla offers sucks.

Their value proposition is that payment systems are hard, so game developers should let them do the hard work and focus on making games. They pull this off by being the merchant of record. They take care of payment integrations and customer service, and take a percentage of the gross revenue on top of the merchant fees.

Their API is not as good as Stripe. And you also depend upon the quality of customer service. When I was Google searching about technical issues with integrating with Xsolla, I keep finding gamers complaining about Xsolla.

If your customer service is better than their’s, there is little point in going with Xsolla, especially if a terrible payment experience will sour any goodwill from fans.

You can fire almost half of your team based on big data analysis, but I doubt it would save your company if your product suck. Customer service people will get demoralized when the issues they hear from the users are not getting fixed by the product and engineering team.

Istm that often enough digital activity is inversely correlated with productivity. Some of my most productive days are spent between walking for a bit, staring at a paper, walking some more, staring at code,... Followed next day writing a few lines of code. How will any tool measure that?

Fire 40% of the people you don’t want and you could lose 40% or more of the people you do want.

One company I worked for decided to drastically change their product direction, meaning no more backend and frontend devs. They fired (made redundant) 50% of devs, devops, dbadmin and some automation testers. 45% left voluntarily in the next 3 months. That was the plan.

So the goal was to keep only 5%?

Yes, one or two developers stayed after 2 years

Consulted at such a place. Entire leadership was eternally gone on "leadership seminars" so entire floors were almost vacant.

> Fire 40% of the people you don’t want and you could lose 40% or more of the people you do want.

Let the market decide.

If this leads to even a 1% increase in profits, expect all companies to follow.

After all, the buy-and-hold index fund investors on HN will fight day-and-night for a 1% increased return for their personal portfolios.

“Tell me you’ve never managed an organization long term without telling me you’ve never managed an organization long term”

Yeah, you initiate significant layoffs in any department, especially in a hot market like this one, for any reason, and you can expect most of your real talent you kept to start looking elsewhere and start resigning, which will lead to more people looking and leaving, which leads to more people looking and leaving, until you've lost most of your good employees. I've seen exactly this happen on a couple of occasions (and stupidly stayed longer than I should have myself).

This is absolutely the typical result, and it may well be the result here. However, I wonder if it is going to change now that more people are working remotely. If you never saw those people in the office anyway, will it bother the typical person as much that they don't see them at work anymore? It may be that "survivor guilt" is less a rational response (if they're laying people off, maybe I should look for somewhere else before they get to me), and more an emotional response to the empty desk next to yours. If you're not even at the desk, you're at home, and it all looks the same as it did before the layoff, maybe the emotional response is not the same.

Or, maybe not. Just a thought. I guess we will find out soon.

"You received this email because my big data team analyzed your activities in Jira, Confluence, Gmail, chats, documents, dashboards and tagged you as unengaged and unproductive employees."

I hope they mean spending too much time in Jira/Confluence/Gmail/chats instead of actually working, but I'm afraid it was the opposite...

I just don't understand why they need to do all this "activity analysis" stuff instead of just looking at whether the employees complete assigned tasks or not.

> we will help you find a good place, where you will earn more and work even less

Wow what an underhanded insult. “We can help you be the lazy, unmotivated slacker you secretly aspire to be”

But if there are jobs paying more and asking less than xSolla then why does anyone work at xSolla?

So you’re saying that anyone can just get any job they want? At any time? Sign me up :)

Many founders are that unusual sort of person who likes working more than anything else, and a lot of them get the idea that this makes them virtuous and everyone else ought to feel the same way.

"Sure, we don't pay as much as some, but look at all this work you get to do!" I've interviewed with people like this. They honestly think that's an incentive (if you're a "good" employee).

It’s the flip side of something that can actually be good, which is not asking people to do something you’re not willing to do, like work weekends for a deadline or take a pay cut.

But your relationship to the company is different than ours. A pay cut still leaves you with a 40% stake in the company, while we get nothing in a fire sale (your preferred stock doesn’t get diluted nearly as much or as frequently as ours does). Equally or perhaps more importantly, this was your dumb idea/baby in the first place, we just showed up to work. You had better be more invested than the rest of us our why are we even here?

I failed an interview once when a founder started bragging about how his wife is sad that he doesn't come home to see her and their young children enough, and he was proud because it meant he was working hard enough. He asked me if I was a hard worker and I told him I preferred to "work smarter not harder" and gave some lines about being efficient and not burning out. I gather that he was not impressed, given I got no offer after appearing to pass the technical screen, getting along great with the engineering team as a whole, and applying to a team where I had direct (uncommon) experience and knew every member of.

I’m that type of person that enjoys working way too much. But as a founder I had unlimited vacation time, which I quickly realized lead to no one taking vacation. So now I say “ten days minimum vacation”, and if you are dragging your feet on taking time off, you’re getting your Gmail, GitHub, etc access cut off on Friday’s, weekends, and or Monday’s when you’ve been working hard.

Currently not a founder, but a senior employee - never mentioned to my CEO that was my policy, and it’s quite nice to have a boss that does the same and forces me into long weekends.

That's a lovely policy. Well done.

I mean he properly leans into it in later comments too, so I don’t think it’s meant to be underhanded, just plainly aggressive

> Wow what an underhanded insult.

Actually, he just signaled why the firings happened. People were claiming to do work, and there was not proof they actually did work in their email, documents or the systems they were supposed to do work in.

Any worker who has carried tons of (usually "rockstar", "bad-ass" or "pedigreed") people who don't even understand the work they are pretending to do (or worse, supervise) knows this is a rampant anti-pattern in business and in many cases has the potential to transform the business. When I was younger and worked for a boss, I used to really hate watching 80% of the others disappear at 4:30PM without regard to customer expectations and come in in the morning, only to start working around 10:30 after making the morning rounds (coffee, chit-chat) the next morning... only to hit a critical lunch with the team at 11:45-1:30pm.

In every company I've managed or started one of the facts I try to be on top of is "actual work done per day". I've found that if I look at productivity and use that to inform who gets invited to help make decisions. It makes a profound difference. People who work hard tend to come to meetings prepared, are respectful of everyone's time and most importantly know the facts about their work. You end up with better decisions.

We saw an antipattern developing at this job where most or all of the people who actually solve problems that come up would agree to use one tool or solution by the greater team would prefer some other solution.

The person who cleans up the mess should get some say in how big and the kind of messes that get made. The rest of you all fuckers disappear when the going gets tough.

I don’t entirely recall how we solved that problem but I believe it involved less transparency, which sucks, but so does burnout.

How do you measure "actual work done per day?"

So, conflicting thoughts:

1) "keep their medical insurance and receive medical pay equal to four to six monthly salaries"...is pretty good. Better than the severance package of most companies in the U.S. However...

2) "we will help you find a good place, where you will earn more and work even less. Sasha will help you get a recommendation, including the one from myself..." is not going to be worth much, since he just very publicly said "these people who we laid off are all slackers". Which was totally unnecessary, not even a mercenary business reason to do that. Looks like a rookie mistake. You should not screw people over, but you especially should not screw people over for no reason that benefits you or anyone else.

They're wanting "busy" employees, which is not necessarily the same as productive ones.

It's like the old "Jesus is coming ... look busy!".

https://www.amazon.com/Jesus-Coming-Look-Busy-T-Shirt/dp/B07...

Dupe (142 comments): https://news.ycombinator.com/item?id=28068959

New SaaS idea: a busyness simulator that sits on your home pc clicking between emails, running semantically correct work-based searches and sending occasional high value Slack responses. I'd imagine the AI ability to do this would be pretty high already.

Beat big data with big data, why not.

I know someone who keeps his msteams saying he is active by using a paper clip jammed into the insert key - that’s probably a good start :)

Watch it just be the employees who went on VK on their work computer instead of using their private phone.

I suspect once people know how these algorithms work, the cat and mouse games begin - making the whole thing pointless - eg the first thing that has come to mind is an arduino set up as a usb keyboard and/ or mouse doing scipted work?

never underestimate people with too much time to hand

Where were the managers? If you have to use company wide data analytics to know whether your employees are productive, you should fire your managers first. Then with better managers, those unproductive employees might start being challenged and engaged appropriately.

I am sure that this kind of analysis has been going on for a long time at some large tech companies. I'd guess that they were smart enough to not announce the methodology. Just like layoffs hitting the pushing 50 crowd, "up or out" mindsets, etc...

> we will help you find a good place, where you will earn more and work even less.

LOL what?

If you were an employee of this company who left, would you resign and look for a new place or would you stay?

Look for a new place then resign

In any order ... question is would you stay in company that measures engagement by the key stroke

I bet the remaining employees' metrics will see a marked increase even as actual productivity drops.

Xsolla is another payment platform for merchants?

Why not fire the CEO for being disengaged from coming up with a novel idea?

The meme is these people work so…much…harder. But none of them built the patterns and infra that enable these companies to spin up in days, on their own.

Without a society capitulating to behave this way, these guys wouldn’t have anything. I see no reason to believe they should be respected for grifting.

1. I’ve built systems that can do this https://insideropinion.com/

2. If anyone has worked in a corporate environment you know a large portion of the team is slacking. That’s not to say they should be fired. I always recommend rehabilitation, as those people already have institutional knowledge AND if they can be rehabilitated are more energetic.

3. If you only now realize their activity, you had bad metrics for success. Activity does not mean results, and results should be the end goal. Case in point, an expert developer may like to take talks to consider a problem. Their activity will be lower, while their delivery will be higher.

4. My system monitors morale as well, because firing someone impacts morale. In this case morale will likely drop, but overall work may improve (I don’t know).

5. In the end, “data analysis” probably didn’t need to be done. People who don’t do work are left alone, if you surveyed the company you’d probably find a lot of those people single out by coworkers as people they don’t want to work with.

Anyway, I always have mixed feelings on this. Personally, I think it’s good to cut weight, but better to focus on making the team fit (and reducing attrition).

Reply on news.ycombinator.com