Atlassian has 16,000 employees worldwide and reported a net loss of $257 million last year. In fact it has not had a single profitable year since its IPO in 2015. And as expected its stock price is in the dumpster, down to $75 from its peak of $440 in 2021 (-83%).
So, like a lot of the tech industry this is simply a case of overhiring, overspending and general mismanagement. And like every other layoff announcement “AI” is a convenient scapegoat to hide executive dysfunction.
This is a pretty surface level analysis; Atlassian also has stock buybacks of billions of dollars each year. It's an intentional choice to not declare a profit and pay a dividend, and instead to reinvest in acquisitions and pay shareholders via stock buybacks.
You'll find it much more interesting to look at metrics like free cashflow, which is a better indicator as to whether the company is generating more cash during operations than it spends. This is a lot closer to the layman's idea of profit, and in a small business like a restaurant or single store it's often analogous to profit. In publicly traded companies, net profit and loss are borderline meaningless.
Atlassian also issues billions worth of RSUs to employees every year. The buybacks are necessary to offset that. Shareholders are getting nothing out of it, and the fact that the stock price has been in freefall for years despite the buybacks shows that.
Software companies hid behind “free cash flow” and “YoY revenue growth” and “non-GAAP profit” for years in the zero interest world. Now that the heat has turned up a notch investors can see what I said before - most of these companies are wildly overstaffed and addicted to spending.
The only reason it’s buying back the stock is because the price has been destroyed by the company’s underperformance. Also yes buybacks will reduce profit but an unprofitable company doing buybacks is not a healthy one.
Agree with the above general assessment that this is a company doing layoffs to address years of bad decisions and underperformance, and the broader issue of SaaS valuations as folks question valuations there. Just another case of “AI” as a PR puff excuse to avoid simply admitting this is about correcting prior bad management decisions.
TIL that 20-30% revenue growth year over year is underperforming.
Atlassian does >$1B/year in free cash flow. the GAAP losses are almost entirely stock-based comp, which is non-cash. the buyback exists specifically to offset that dilution.
> buybacks will reduce profit
wrong. its a balance sheet transaction. cash goes down, shares go down.
There is no connection between profits and dividends, some monstrously profitable companies pay little to no dividends and return money to shareholders via buybacks. The act of doing an acquisition by itself also does not affect profits.
The reason Atlassian doesn't "declare a profit" is that they are not profitable - they pay a lot with SBC, thus diluting shareholders year after year and not returning anything.
Profit is just one metric that affects the stock. YoY Revenue growth was historically in 40s and high 30% year on year. But last year was sub 20%. Investors are re-evaluation SaaS company multiples through a new lens of AI. Personally I wonder if Fortune 500 can renegotiate lower margin pricing with the "We could just build this with AI in house" BATNA argument.
> In fact it has not had a single profitable year since its IPO in 2015.
What's frustrating is that the company was profitable from its second year, and a really good little money maker too. I was working there at the time and I was proud to be working at one of the few tech companies actually turning a profit. Then IPO came, then literally overnight we switched gears and "reinvested every dollar into growth" and decided we'd just be another dumb money losing tech company.
Your assessment is spot-on! These entities have been on shaky ground for a while, and AI is a convenient scapegoat for justifying their unavoidable downsizing.
Atlassian was famously bootstrapped and profitable since inception. They were one of the only profitable tech companies at a time when everyone else was busy making bigger and bigger losses.
If they have really been unprofitable for the last decade that is very sad.
Atlassian is cutting jobs because no new sane company wants to use their products. Confluence was once innovative but now has gone stale. Jira is a nightmare and is most ripe for the AI based replacement. Bitbucket is a neglected product that has lost ground to GitHub and GitLab. The writing has been on the wall for Atlassian for years.
The peak was right before they launched marketplace in my opinion (2013?). After that they had no incentive to improve Jira/Confluence, easier to take the marketplace cut. I could never understand why they couldn’t solve the speed issues though. It just got worse and worse.
I tested out Atlassian Rovo last year. I tried to get it to list all of the Confluence articles I had written in 2025 so I could use that information for my performance review. It found three, regardless of how I queried it. I had actually written over sixty. I tried, but never did found a good use case for it. Too unreliable.
I tried to pin my comment. It turned out after we migrated to the cloud version, this option disappeared and the ticket is flooded with commit messages making it essentially useless for humans. Rovo obviously faked a solution:
Atlassian Rovo is considered utterly dangerous on my team (as we are still encouraged to use it…), the first time I did it erased a whole page - it couldn’t get the most basic instruction correct, would leave out 80% of what I input, respond to corrections with the same problem. It’s just a liability.
i agree that their products are quite stale, but a lot of that is cuz they're so deeply entrenched in enterprise workflows in a way that's incredibly hard to displace. switching costs for jira alone across a large org are brutal.
could an AI native competitor eventually eat their lunch? sure. but "no new sane company wants to use their products" is a stretch when their customer count is literally still growing double digits.
I just started using Linear, it is AI native/friendly and a million effing times less bloated than Jira. So long Jira. And I say that as an old shit who migrated from Bugzilla to Jira a long time ago and I remember how refreshing Jira felt at the time.
As digital post-it notes, I’m not sure how I would replace Trello with AI. But otherwise I think your going to get mass agreement that Atlassian is running into the ground.
It went cloud-only and Atlassian discontinued the self-hosted version of Bitbucket. Not strictly neglect, likely just a business decision, but even still the cloud version of bitbucket remains a subpar UX.
I see. So AI is reducing the number of jobs in the tech sector because fewer people are needed to ship stuff (thanks to AI). And since fewer people are needed across the tech sector then we don't need things like Jira anymore because it can all be done on post-its or Google sheets or something, so there's no need for Atlassian accounts anymore. And Atlassian can now do more with less thanks to AI.
I can't wait for Atlassian physical sticky-notes to take over.
Imagine doing AI development in waterfall. You spend weeks writing your prompt, when you think you have it perfect, only then do you submit it to the AI. Then you wait a week or so, and see what it produced, expecting it to be exactly what you wrote.
Or, do you tell it the basic functionality you want, test it out, then add feature after feature that you want, sometimes dropping them and sometimes adding new ones that you thought of as your worked.
I don’t think that’s what it would look like at all. The first stages would be cheap - mostly requirements gathering and research, but a bit more focus on that. A bit more time would be spent up front, but then you’d see multiple proposals being built, from
that multiple plans being built, and finally multiple implementations to choose from. You might see A/B testing of multiple implementations or even products, and then a decision on which to pursue. You could move in multiple dimensions concurrently.
I’m not sure this is agile. I’m not sure it’s a waterfall.
We’ve got bounds on our infinite typing monkeys, but they increase every day
> Or, do you tell it the basic functionality you want, test it out, then add feature after feature that you want, sometimes dropping them and sometimes adding new ones that you thought of as your worked
the problem with this is long term maintainability. it works - and the engineer understands how it works - but a) the AI does not prioritize cleanup/organization/naming, and b) there's a blind spot/boiling frog type of phenomenon that can prevent the engineer from spotting the growing problem. the codebase becomes recognizable only to them. the engineer sees all features working, all bugs fixed, 90% test coverage, and submits it for a PR.
the engineer tasked with reviewing the PR will treat it as slop.
The vast majority of "AI is changing everything!" takes I read say more about people's fundamental misunderstandings of the software development lifecycle (the real one that companies actually do, not the one that people think they do or what companies say they do) than about anything AI is going to change about software eng.
If anything, their solving the complete wrong problems and being blind to the actual problems is probably a reason AI won't actually result in any real, top-level appreciable gains in shipping speed.
Waterfall came out when hardware and software had to be developed together, and appealed to traditional Engineering practitioners. You are right though, when the hardware constraints went away, software (more code) was cheaper and easier to ship in increments and iterate. But feature-rich products were still difficult to ship - and you had to pick and choose what things to spend your time on.
The SaaS-pocalypse is occurring versus investors don’t believe that to be true anymore.
I think they will still be wrong because ultimately people want people (particularly experts) to be held accountable for things - shipping high stakes software, running company ERPs/CRMs, and more.
Honestly posts like theirs are just indicative of someone who never understood their job/role.
People throw out terms like agile or waterfall, shit on agile etc. probably because they work at some worse than mediocre place let alone ever done their own thing.
Glad you mentioned Google Sheets. I moved my personal task tracking from Trello to Notion to Sheets. Sheets has been the best for me. Infinite customizability, fast, lean.
I'm curious how you use Sheets as a Trello-like replacement. I use Trello from time to time but my most recent project is using actual Post-It notes stuck to my monitor, wall etc.
If Sheets works (I love simplicity) then I'll give it a go but I checked and I can't find any templates for Kanban or such.
when agile was fairly new I worked with remote developers that had 3 locations.
My specialty is software requirements and my team was brought in to do the product management. The developers had read somewhere if you were using a database to do requirements then you were doing agile wrong.
They wanted me to write post it notes in triplicate, then fedex them to all their offices.
I'm pretty confident though with no solid evidence if you lower the first number by 1 you are describing the vast majority of employed programmers in the world.
I don't think it's completely right. Jira is a task manager, and the task throughput supposedly remains the same, just fewer assignees.
I think these companies should be pivoting to something where tasks/issues are the places you write the prompts for the AI, or augment the prompts that devs use. It's a big shift though.
In my last jobs Jira was used, and despised by all except product managers. It just becomes a mess.
In my startup (now 20 people), we use Trello. Outsiders look at us funny. I respond that its the same company after all...
That's because these layoffs aren't about AI. They're about firms that overhired and Wall St is (finally) having a sobering moment of their (profit) growth potential.
On a more serious note, at this rate, probably 10 years. I guess it's similar to drinking. You can get drunk in 10 minutes and be hung over for much longer than that.
It turns out that org charts are VERY resistant to letting people go, even when the executives push very hard.
> are betting that their AI business will be profitable
It could be. The idea that a business built on this framework won't immediately be picked apart by every competitor under the sun is entirely beyond me.
> and need to cut costs to invest in it.
If I'm right then this is just slashing your own throat so you can be the first to the bottom.
> They are betting that their AI business will be profitable, and need to cut costs to invest in it.
I think AI is a wonderful excuse for many. You can now lay off as many people as you want without standing out in a negative way. "Everyone is doing it" and "everyone knows why it's happening".
It makes quite a bit of sense if the size of your market doesn't expand along with the new technology and you don't have a competitive advantage. Just because you have the capability to deliver more packages doesn't necessarily mean you'll have customers willing to pay you to deliver more packages.
Can someone fix the damn Slack plugin so it stops asking me to turn it on every time someone posts a Jira link? I have dozens of messages from the plugin saying "Got it! Will not ask you again."
Right! There is more performance to be taken away yet! They're nearly there but plenty is left. Jira can be made "let's go brew coffee and come back" slow if the really tried
You have a package delivery business. The business is stable and optimized, serving all houses in your city. Each courier has a route planned out that they follow each day. They take 8 hours to complete the route and deliver to each house along the way.
Now, cars come along. With the new efficiencies, you find that now a courier takes only 6 hours to do their route. The number of buildings in the city has not increased, there's a cap to how much service you need to provide after which there's no longer any additional benefits. So, do you cut everyone's work day to 6 hours, or do you fire 25% of all couriers and re-route the rest so they now do 8 hour days on longer routes using cars?
Your solution assumes that the extra workforce immediately translates into free growth, but the growth of many businesses is constrained by outside factors. Here it would be the population of your served area.
You need a very different skill set and culture for driving and maintaining cars than for driving and maintaining horses. I honestly think that for big changes like this (if you are willing to accept that AI is such), looking at it from the portfolio management angle, it makes more sense to just nuke the current operation and start a greenfield one.
Senior engineers have been 'vibe coding' for over a decade before AI. Think what they do, they look at PRs all day and comment. Magically the code change reflecting their comments. It's the same thing now but machines are doing it, not humans. The issue is that junior engineers have no experience working like senior engineers. The reality is that it's not that hard to work in this way. There is no excuse for software companies not being able to re-train their more junior engineers to work this way.
That's a very nice analogy. I agree and thinking about my previous comment, I suppose I just lashed out because I really dislike Jira the product and don't think that it can be salvaged, but I don't have anything against the engineers working there, and agree that they can be mentored and reassigned to a product where they'd be able to create something good.
No, the assumption is that companies are more interested in cutting labor costs than productivity. Even if you screw up and need to hire back 50% of those you fired, you still cut the labor costs of the 50% still fired. And you can pretend to be a cool, thought-leading, "AI-native" company, which might be enough to juice your share price enough to offset any actual productivity loss.
Capital will always be in opposition to the cost of labor and want to make it as close to zero as possible, and AI is a plausible story for attempting that, regardless of the reality of AI efficiency.
It is amazing how that bit of corporate PR is still being quoted over 100 years later. In reality, Ford had huge turnover problems with his workers - one estimate is over 370% annual turnover. One way to help prevent turnover is to pay more, and it solved the problem. (Even so, the base pay was still actually $2.30 and to get the extra $2.70 you had to abstain from alcohol, keep your home clean, etc.)
The strategy for institutional investors is to invest in servicing the needs of the already rich, at the expense of investing in companies that serve working people. The former is much more profitable than the latter, and the latter is becoming less profitable over time.
No the assumption behind the statement and many others on this topic is not AI is more productive than Human + AI, it’s that 9 (or 8 or 6) humans + AI is more productive than 10 humans. No one is suggesting getting rid of all workers, but many are saying they can get rid of a significant percentage of them. It remains to be seen if that ends up being true but it is fundamentally different from what you are describing
Then you are not laying off because of cars. You are laying off because there is no market for what you are selling and you are not able to find new customers.
I'm not saying Atlassian is or isn't doing well, but your market isn't going to infinitely expand, nor expand at the rate you need to keep up with your costs.
If you build cars and you replace people with robots, you might fire those people because those robots allow you to keep up with demand with less people. Your business is doing fine and well but you just need less people to do the same output. Better yet those robots might even produce more than what your demand is and therefore even if demand increases your ability to supply with less people still exists.
Businesses will always look to do the same output or more with less cost irrespective of whether the company is doing well or not. This is why we have also seen "AI layoffs" across companies that have had very good financial positions or even record sales.
The proof is in the pudding if AI actually improves productivity of those left over to justify it. But sometimes simple analogies can be rebutted by other simple analogies.
> your market isn't going to infinitely expand, nor expand at the rate you need to keep up with your costs.
The costs don't change. They were already paying those people. They just need to take advantage of the slight bump in productivity LLMs provide, not infinitely expand their market.
> Better yet those robots might even produce more than what your demand is
You are repeating the same idea from your original comment. The problem is that the demand is low. You are making something no one wants to buy.
> Businesses will always look to do the same output or more with less cost
this is not an accurate in general, but especially not in this situation.
There is a magic box that makes all of your employees get 50% more work done. You are an early adopter of the magic box in your industry, though others will likely follow soon.
Do you A) do 50% more work for free and get an early leg up or B) fire staff until you are achieving the same output as before, at a slightly lower cost.
Which seems the more logical option in the long term?
The only reason option B makes sense is if business is not going well and you can't afford to keep paying everyone.
Chartr Daily had this chart [0] back in 2023 and it shows how much the big tech firms grew from 2016 to 2022.
Some of the firms, Apple being the exception, doubled or even almost tripled in size.
I'm sure AI is partly to blame here but I think a lot of it is over hiring and firms just getting bogged down in bureaucracy and trying to clear things out.
I don't think AI is even partially to blame. Unless Atlassian is claiming AI can fully replace 1,600 workers, layoffs don't make sense.
You need people driving AI to get the benefits.
Its like a courier service that uses horses firing people once cars are invented because cars are faster than horses. You would switch everyone from horses to cars and deliver more packages.
Well you might not, if your accessible market is 10 horses and 5 cars can fill that market need, then you're left with 5 people who aren't needed because your products don't meet the needs of 10 cars worth.
If your postal service services a population of a million people and it takes 1000 horses to do that easily, but it takes 500 cars, you don't have a need for those 500 extra people.
You can't deliver more packages if the packages aren't there to be delivered. Your product demand doesn't just magically scale up once supply meets it.
Good point, but what if you were previously chaining horse carriage rides and now a car can cover the same distance as 10 of them with a single driver?
I don't know, maximum package turnover might be bounded and most likely you were previously not constrained by lack of drivers already... Sure you might try and expand but why would that work better than before? Especially assuming all other providers now also have cars.
If they are paying them 6-month severances like Block did, this means they are effectively saying 1,600 people for 6-months wouldn't have fixed JIRA's usability and performance, which if they could have done like many have been begging, they'd would probably make more money long-term than this firing would save.
I wholeheartedly believe that they could not have fixed it with 9,600 people months of work. They haven’t been able to fix it with many multiples of that.
> They haven’t been able to fix it with many multiples of that.
which may actually be the problem. I suspect that there is actually some ideal ratio that could be calculated of Input Fields / Dev, LoC / Devs, or maybe Unique Pages / Dev, or some mix of all of the above. Some of the metrics I hear out of places like airbnb absolutely blow my mind (>5000 engineers! wtaf are they all doing?!?). I can sort of see the #s at google, MS making sense given the breadth of the problems they are solving, but other places, not so much.
There was an interview with a lead tech at Uber about this some years ago, with the conversation starting with "why is the app so bloated!?" (in terms of megabytes) and his answer also answers your question:
The smooth and simple interface of the Uber app is the tip of the iceberg. His example was that their users don't (and won't) reinstall the app because they travel overseas. If anything, the time when you've just stepped out of an airport is precisely when you want the app to work smoothly!
The hiccup is that many countries have their own payment systems, Byzantine tax codes where this may or may not be displayed up front to the user (in various currencies and formats), there may be local laws around taxi-like services, etc... Some of those laws apply to areas smaller than a city, or may apply only to airport pickups, or the CBD area during congestion, so on and so forth.
The "core" app might be a simple thing that you can bang out over a weekend with an AI and a decent UI framework, but then you need to "draw the rest of the owl". Don't forget that there must be a matching app for the drivers! Different categories of drivers offering services that may be local to a region and totally absent elsewhere: rikshaws, tuk-tuks, taxi boats in Venice, and who knows what else!
AirBnB is very similar to Uber in this respect. They have to deal with about a hundred countries worth of law, often down to the state level. There's fraud detection. Customer support. Integration with travel agencies. Government-mandated reporting. Etc, etc...
You're assuming performance has been the core priority, or even a priority at all, and I think this is a bad assumption to make. I would estimate a much smaller number of people-months of work if I were you.
Dev users assume the only problem a product can solve is performance, when there is a lot more than that in reality.
Maybe in the past companies wouldn’t take the extra time for performance enhancements - but they’re apparently saying that AI is sooo good and speeds up work that they don’t need all of these extra people. So if their product was sped up it would enable their customers to work faster and lay off all of their extra employees (or just keep everyone and just do more stuff faster).
So are they doing this to make the product better or, as others have mentioned, they can’t innovate further and can’t grow their market so they need to cut costs.
They need to just clean slate start from skratch. I don't believe that code base can be saved.
AI means it's easy to copy any SAAS now right? so should be easy /s
I mean probably not. It's not 1,400 hardcore engineers and 200 Jony Ives being let go, it's a mix of everyone including randoms like HR and the person who orders the office coffee. Business is not good.
Okay I just wrote an "it's not, it's..." organically, is this the zeitgeist or what.
It's not a zeitgeist, it's a common English sentence structure.
The em-dash observation makes sense. Obviously a minority of people reach for an actual em-dash. The "it's not X, it's Y" complaint is totally bonkers. We might as well call proper spelling a "sign of AI". I.e. yes, AI does it more than humans, but not by so much that it makes sense to be suspicious when you see it.
We use Jira. We're trapped by the processes built around it. Once it's entrenched, it's rather difficult to remove. I eagerly await for something that can replace Jira, but fear that it too will be bastardized to fit the process for however long the process lives.
Linear has a very different vibe to it, but that's not inherently a bad thing. Have used Jira plenty, and even ServiceNow Agile Development (FML, makes Jira look like a paragon of software experience), but Linear is quite refreshing.
If someone like OpenAI or Anthropic pulls it off, and imposes strong opinions similar to SAP (your business adheres to their vertical model vs them tailoring SAP to your unique business), I think it could replace Atlassian tools (Jira and Confluence specifically) relatively quickly. Call it “Planner” or something similar. Tell the Robot what you want and have it build and manage the plan. Atlassian’s revenue is their opportunity.
The cloud version I used was slow and riddled with bugs. Entire views sometimes just refused to load or render, or something.
Did it "get the job done?" Yes, in a literal reading of those words, I suppose it did, but anyone who understands the amount of work that a modern 2.4 GHz CPU should be able to do per unit time would not think highly of it.
Nowadays … my company uses Linear … which, while it does have a sleeker, more modern looking UI … is nobody able to make a good bug tracker?
I haven't looked at it recently but always felt that tools like Jira encourage poor management. Effective teams are small teams, and small teams should be communicating and working together across issues freely and frequently. It's generally harmful to have a manager assigning tasks outside of the actual day-to-day discussion without having to speak to someone directly or preferably in an open chat or thread where people can see the discussion.
And ideally the user facing chats and threads are directly linked in to the development chats or at least with channel notifications somewhere.
Task assignment tooling encourages managers to stress developers out with low priority tasks that often start off with incorrect requirements that the structure makes it hard to correct because it is then directly a disagreement with your boss and there is inherently not a discussion it. Whereas a chat at least has the concept of an informational response to a nonsensical task as being fairly standard.
> Its not slow like the on-prem. Jira cloud version is fine.
Seems like opposite land to me. Back in the day running Jira Server was the only way to get a snappy Jira instance. When they discontinued Jira Server to force everyone to the cloud it was god awful slow and forced us to abandon not just Jira but our entire Atlassisn stack.
I don't 'like' Jira, but it gets the job done. It's so easy to onboard users and assign tasks/issues across orgs. Structure is fairly simply and the filters with subscriptions is powerful. Android app that I use on my work phone just works.
No, there was a big internal project (which they communicated publicly about - search for the blogs relating to it) to address it that involved roughly a year of effort from a big chunk of the developers.
Linear is great if you fit into its workflow. It's very dev orientated.
I work on Aha! Develop https://www.aha.io/develop/overview which I obviously think is a great tool, especially if you're a team with a product manager.
That was the wrong word. Something about the UI for it feels quite constraining, and that isn't necessarily a bad thing once you're used to it, but at the start of using it I struggled to get what I felt was a broader view of how a project was going.
I've only used it on a side project though because I really wanted to get to know it. I know plenty of teams really love it.
Have you tried nothing at all? Had great success with this on a 150+ dev team. Much preferred to jira. Admittedly does require a different approach to work than a jira-centric team is going to be familiar with.
I 100%, sincerely agree that "nothing at all" is really a great option compared to Jira. I actually hired a whole person onto my project to make them responsible for ticket tracking, telling them they could use whatever they wanted so long as I never had to look at Jira again. They used Jira for a while because it's what other people were pressuring them to use, but ultimately they started using MS Planner. MS Planner! I mean, Planner is garbage too, but at least it's not Jira.
I do not understand why Jetbrains Youtrack is not more widely used. It's affordable, supports markdown, has a better vcs integration, fast & easy API...
I like Clickup's way of allowing arbitrarily nested subtasks and easily promoting/demoting a task across levels, without having this hard distinction that Jira has between levels. I understand that some coporate managers like the rigidity, but in practice, it's just very hard to know the scope of a story early on, and I found this flexibility really valuable.
Thanks, but just to confirm, there is a free tier of OpenProject as well as a paid tier with more features. This is often referred to as "open core."
I'm not complaining, as the free tier seems feature rich by itself. Just wanted others to know, especially if this has a chance of changing over time once people have committed to using it.
I don't know if it's "popular", but I use Clickup at work and I think it's generally fine. At least when I have used it it's less laggy and horrible than Jira.
The problem is "no one was ever fired for choosing Jira" (or Confluence). I can't imagine there is a single company that would keep using those products the moment some alternative exists that managers dare to switch to without the slightest risk of having to explain to their managers why they didn't just go with the "industry standards".
Second this. Funnily, their AI is pretty good... recently I accidentally created a dupe ticket; linear AI realized that, labeled it as a dupe instantly after I created an issue. Fuck JIRA and confluence to hell.
I don’t think anyone believes this is because they are becoming more efficient because of AI. It may be a bit because AI makes their products even less attractive than they already were.
> The authors find that height cannot, in fact, be used to predict changes in GDP. However, GDP can be used to predict changes in height. In other words, the study finds that extreme height is driven by rapid economic growth, but that height cannot be used as an indicator of recessions
While we’re at it, can Jira have a Slack integration that takes a single emoji to create a ticket like Linear does, instead of a four click menu driven flow?
Please don't tell me that Jira is about to get even worse...
I don't understand the AI layoffs; there's always an infinite supply of new work that could be done. Instead of firing 1600 people, why not have all of them use AI to produce more stuff and outrun their competitors.
Presumably all their competitors also know about Claude as well, and a lot of these 1600 people will go work for them and use Claude.
Unless this is just regular layoffs, but they know if they brand it as "AI" their investors will eat it up.
If everyone else is downsizing and using AI as an excuse, it's both a pretty good cover for any firing you might have wanted to do for a while, and you can reasonably assume you can hire back in the future because everyone else is firing too. Maybe you can even depress their wages a little?
I think this is exactly what/how C-suite thinks. It's incredibly dumb because of how much hiring/firing costs in terms of ramp up/context/productivity etc, and inevitably, they will have fired people that are great and have kept people that suck but know how to play the game.
> there's always an infinite supply of new work that could be done
I distinctly remember a discussion where someone says "Man, I wish JIRA would add this feature/fix this bug"
Someone else pipes in: "I bet there is already a ticket on the JIRA bugtracker/feature board for this, it's not done and it's from 9 years ago" and lo and behold there was.
Unfortunately, none of these companies are going to turn their AI loose on important, annoying, 9 year old bugs. They're just going to use it to cram more unwanted features into their software, just like they're doing today with human developers.
It is just regular layoffs, and doing so admits they don't know what to do with the 1,600 people anyway, and probably didn't know what to do with them for years.
AI isn't going to help, but it bandaids over the issue so the investors aren't spooked.
Laying thousands of people off often implies you hired thousands of more people than you actually needed, which makes investors feel like you're wasting their money. If you say "no they're all being replaced for $200/month of Claude Code!" then it makes you look like there was actually strategy to this.
> there's always an infinite supply of new work that could be done
I definitely buy this for the software sector or the economy as a whole, but for an individual company? Seems one would be bottlenecked by various factors quickly.
Perhaps better to let people go so that they can be productive elsewhere?
There's always bugs that can be fixed, there's always optimizations that can be done, there's always a feature that someone wants to build but hasn't had budget to do. There's always improvements that can be done for deployment. There's always ways of reducing memory. There's always ways of reducing ongoing expenses etc.
I have worked for a bunch of companies, and even relatively new and young companies have all these things pile up pretty quickly.
Have you tried looking for a job recently? The job market is cooked and it's not getting better any time soon. The supply of candidates is way up. Salaries are going down. Even mediocre jobs show 100+ applicants on LinkedIn.
> Perhaps better to let people go so that they can be productive elsewhere?
True. Joining thousands of other unemployed developers sending applications into a job posting for a nonexistent role online is very productive. Probably good for the economy too now that I think about it.
> Instead of firing 1600 people, why not have all of them use AI to produce more stuff and outrun their competitors.
Alternative take: I can't speak for BitBucket because I've never used it, but I've had enough time with JIRA and Confluence to last a lifetime, and these products are so bad - so clunky, so slow, so much friction in the UI - that I can't really see what useful value adding work Atlassian's 16,000 employees have actually been delivering. From that perspective losing 1600 of them seems like it's not likely to make much difference since, from my perspective as a user, they didn't appear to be doing anything useful in the first place.
I'm sorry if that comes across as a particularly savage take but Atlassian have wilfully been churning out absolute garbage for at least 15 years now (there was a time, in around 2006/7, when I thought JIRA was quite good - genuinely) and their products have made me miserable throughout a good chunk of my career, so my sympathy is pretty limited. If they can be bothered to make the products better, faster, more usable, and remove friction ruthlessly at every turn in their workflows, then I might well change my point of view.
I last used Bitbucket in 2021-ish. It was fine for what it was, especially on-premise, but it's a very hard sell in a world where GitLab and webhooks exist and even harder now that Gitea is finally picking up steam.
I haven't used BitBucket in awhile but I remember it being "not that bad".
I agree with pretty much everything you said; I don't actually think that it's due to AI is my point. If their products are terrible and they're finally losing business over it, it makes enough sense to fire 10% of the workforce. I just don't think AI has much to do with it.
It's likely not all this, but i expect an element is: there is a meaningful number of people essentially refusing to work with AI.
Antidotal but I have spoken to friends at Google who are telling me many co-workers say "I tried it didn't work, ill do it myself" when really they just didn't try very hard at all.
Edit: that is to say, if you had a % of your workforce avoiding helping you explore a current trend (valuable or not tbd sure), I can see rational arguments around removing them from the team.
If, as a member of the c-suite, I find that a noticeable percentage of my company's workforce isn't "helping to explorer a current trend" then either they know something I don't or I haven't given them the time/methods by which to explore.
The latter is actually the more pertinent item, as I've seen several times that an initiative get rolled out by leadership, some teams have free time to play around and use it, and other teams have so much on their plate that they're barely able to keep their heads above water, let alone take on another experiment. If someone is worried about getting a project knocked out by end of month/quarter/year in order to keep their job, they're not going to mess about.
Now, that's a leadership failure, but it happens more often than not.
To add to the speculation, it's possible that the people refusing to use it are working slower. Even if the code that they write is objectively better by any metric you'd like, humans can't really pump out code as fast as Claude or Codex can.
If you can get something into "good enough" territory in 1/10th the time of someone who can get it into "great" territory, that is often worth it.
This is a tacit acceptance that AI maybe isn't as great as they make it out to be.
And since the fact that claude code is an electron app and not AI generated optimized binary per platform, it's abundantly clear that perhaps AI is not all they hype it up to be.
The comments hit at some, but not all, of the underlying drivers. I'll add a more comprehensive view and let you draw your own conclusions:
* Their balance sheet paints a messy picture. Their gross profit per quarter doubled from 23Q1 ($668mn) to 26Q2 ($1.35bn), but their net income has been a consistent loss - from -$13mn in 23Q1, to -$42.6mn in 26Q2. The company has generally failed to turn a meaningful profit after considering operating expenses, reflecting misaligned priorities of leadership.
* Their headcount similarly whipsaws of late. In 2021, it was 8.8k; by 2025, it was 13.8k; in the middle of COVID, it was as low as 6.4k. Even after these job cuts, their headcount remains roughly flat from 2025.
* Cutting jobs to invest in AI when you're already slowly bleeding cash isn't exactly a winning strategy. Atlassian's products have the benefit of organizational "stickiness", and their push to a cloud-only SaaS model hasn't gone all that well if you read the IT rags (lots of uniquely complicated migrations that don't transition well 1:1 to SaaS).
* That said, pointing to AI while cutting jobs isn't a bad play when you're courting investors, many of whom doubt the long-term viability of the XaaS model when AI can slop up boilerplate and internal-only solutions on the fly. If they're doing it to genuinely cut costs and try and right the ship, fingering AI isn't a bad cover.
* Except the reality is most of Atlassian's leadership gets their comp in equity, which has taken a serious hit of late on the markets just as vesting schedules wind down and leadership is changing over. I'd be on the lookout for SEC Form 4's from insiders in the coming weeks to confirm whether or not this was the case.
The reality is that the "AI layoffs" ploy is almost exclusively a cover story for corporations reasserting dominance and power over workers after a few (comparatively) good years (WFH, higher pay increases, wage gains, flex-time, etc). Every single one of these entities obviously has more work than people to do it, but if they can squeeze 90% of the workforce for 110% of the hours, that's a net gain for the corporation and a net loss for workers.
Efficiency, over-hiring, right-sizing, AI; it's all bullshit smokescreens for greed, plain and simple. Don't be fooled by narratives to the contrary.
It’s not that their employees are no longer needed, it’s that their product (jira) is no longer needed. When you’ve got AI agents taking bigger and bigger steps, you don’t need to micromanage people through jira as much anymore. Companies will likely switch to something lighter.
Jira regularly makes it to the top of lists of the most hated enterprise software, there’s definitely appetite in the market for a replacement.
This is developer wishcasting, to be frank. AI has not obviated the need for Jira and the idea that companies are moving to "something lighter" (what are they moving to?) has no basis in reality.
Maybe you’re right, maybe you’re not. But if you’re right i think it’s more “investor wishcasting” than developers.
It really doesn’t matter what us devs think. Investors and industry leaders have decided that AI development is the way forward and we’re going to be managing teams of agents from now on. So we’re not going back to fine-grained task management in jira - what used to live in jira will now live markdown files, and largely be written and read by agents.
Higher level tasks might go into something like Linear, who knows.
If the investors are wrong, and this is all fantasy, then maybe people will go back to Jira, and Atlassian stocks will recover.
If I was controlling a swarm of 1000 agents who do all the dev work on my extremely complex enterprise SaaS product with tightly defined business logic I feel like I’d definitely want something like Jira to manage what they’re working on.
They could also pivot to developing something for that exact use case if it’s a bad fit and if there’s such clear demand.
But truly they’re just cost cutting here and AI is really neither here nor there.
I don’t know about established companies pivoting but new operations/projects don’t seem to default to Jira like they did previously. In my very non-scientific sample size, I’ve noticed a shift in the last 3-6 months
What is everyone shifting to, and at what scale? I can see moving to breaking down tasks in separate markdown docs for a small(ish) startup, but working at a company of more then say 1k or so requires a bit more infrastructure to deal with the cross cutting concerns (compliance/legal, pm's, leadership, etc). I'm at a reasonably sized F500 and Jira is the default, despite how much all of us despise it, mainly because it ticks all of the boxes for aforementioned areas.
Yea, I’m just consulting with small projects. A year ago Jira felt like the default, last month I learned about Linear.
I’ve just been surprised lately at how often it isn’t the first thing mentioned. I came from a Fortune 500 and Jira was mostly hated there as well. I still think its foothold is so big that it’ll be a long time before anyone goes through the effort to migrate. I’d personally rather live with it than try to replace it.
I mean, if they are using more AI and less of the devs who made it what it is... it might be better? A little tongue-in-cheek, but I find jira and confluence much less annoying now that I just made a claude skill for each of them and now I don't have to interact with their UI very often anymore
The 1,600 number blew my mind, and then I realized Atlassian employs almost 16,000. Kind of crazy.
Either way, I'd expect that those 1,600 people using AI to solve Atlassian's big problems would be better for the company in the long-run than reducing headcount with the same level of output
One time charges are pretty typical when layoffs are announced. They are usually the cost of severance pay for the weeks or months of salary paid to employees who are no longer working. Office space leases are typically long term (multiple years) and accounting rules require they recognize the expected future cost of that now-useless space when the layoff decision is made. In practice, cash won't actually change hands for the office space until rent is due in future months. And companies will work with the landlord to get out of the lease (but often pay some penalty for the privilege).
I’d guess some of that is accounting for taxes on profit they’ll now have to pay that otherwise would have been deductible if spent on salaries for R&D?
Also, please note that maybe not all employees on this layoff are US-based.
In several countries, laying off people come with legal requirements for mandatory minimal severance, health insurance extensions, legal taxes and government fees and all kind of compensatory one-time payments for the fired employeer.
Atlassian is cutting another 1600 jobs because it needs to cut more jobs as it is a dying company with terrible products.
But let's try to spin it up as if we were some kind of AI mavens who are reaping humongous increases in productivity due to our thought leadership in AI.
Atlassian pretending they can pivot into AI, is the most "Hello fellow kids" corporate moment this year.
Their services are barely usable with extreme bloat and lag. With such strong engineering practices, they are poised to make fools of themselves. Can't wait.
They started nagging every user in Jira to use their AI, now. It’s like straight out of the Microsoft “Dear god please please use our AI product!” playbook.
I’m honestly not sure what you even use AI for in Jira. Maybe there’s a purpose, but 90% of us are just moving tickets across the most expensive kanban board that money can… rent.
Hence the comment you replied to saying "they print cash". You'll find a lot of big companies work this way: high free cashflow, because they earn a lot more than they spend, but then the spare cash is either reinvested or paid out via share buybacks. Declaring a profit isn't advantageous compared to the other options available.
Anything in particular? I first used it about 10 years ago, on prem, and am currently using the cloud version. Current edition is clunky, slow, and constantly badgers me with Rovo shit I can't disable. IMHO, it reeks of a product once built by and for technical people that eventually got dumbed down by POs to the point of being painful for the original users. Obviously I'm no longer the target user because I assume someone somewhere appreciates these changes.
The performance, at least on our cloud instance, has greatly improved. There was a period of time where every load was causing the side bar and other components to slowly load with skeletons everywhere.
There are still skeletons but a lot of the page components load instantly(cached) and others load quite quickly.
Tons of UI "jank" has been cleaned up across Jira and Confluence. The UI design, in general, has also finally cleaned up nicely and "settled".
Confluence articles load very fast now. They have also added Live documents which is a very welcome addition.
Is Jira still bloated? Yeah, but that doesn't preclude improvements. It feels less bloated now to me.
I'd argue the opposite. The thing is so bloated and the simplest of things seem to be so hard. Import markdown in confluence? Nope not natively. Add an issue to the board? better go to the one workflow to do it and not in the actual ticket.
Confluence used to be built on top of pretty standard plain wiki markup that could be edited without being forced into a bad visual editor, even easy to edit in an external text editor to not have to spend so much time in the web UI at all. I remember having an Emacs mode for it installed.
Looking this up now, Wikipedia says the wiki markup was abandoned already in 2011. Not that I think Confluence was ever a great wiki, but at least having pages that were backed by some resonable plaintext markup was much better than not having that.
Perhaps I just haven't noticed them, which is unfortunate. I have noticed that I often have to double or triple click to open a ticket on my board. There's no reason for such a core functionality to be that slow.
The only Jira improvement that I can remember from the last 11 years is being able to drag and drop subtasks (maybe it was up and down arrows before that).
Everything else has been UI changes as far as I've seen.
Same! This nightmare seems to have started recently from what I have experienced. I find myself hard refreshing constantly to get the damn tickets to open.
If anyone from atlassian reads this: please, for the love of god resolve this issue.
I honestly hope someone will read this comment and vibecode an Atlassian 2.0 platform, preferably open source. But really, I will take closed source and paid as well - just give me something that's on par in terms of features and integration but without the terrible UX.
To be clear, I agree with the terrible products part - but currently they are not dying because there is no alternative platform which is flexible, scalable and feature-complete enough. You may find alternatives for niches, like GitHub for software engineering, but the Atlassian stuff allows for knowledge transfer and familiarity across many many domains. I've seen it used anywhere from government burocracy to customer service and construction companies. They nailed the abstraction for flexible issue management, just the implementation is terrible.
I vibe coded a native client for Jira that’s speedy for creating tickets. At this level, you could write something native and just use their API and have it be as quick as you’d like.
Its weird to hear Atlassian being associated with productivity boosts of any kind considering the zillions of developer hours wasted trying to navigate JIRA.
Seeing this kind of news makes me very uneasy. There's no doubt that AI is more efficient, but what about diversity? If many companies prioritize AI over humans, won't society as a whole become very uniform?
Engineering including CTO. The makeup of Australian staff cannot be immediately decided due to local workplace rights laws requiring genuine consultation.
> The union representing Atlassian workers, Professionals Australia, said impacted employees were told on Thursday, with a consultation process to last until 19 March and final termination expected on 2 April.
> More than 900 affected positions were involved in software research and development, a spokesperson said. Most of Atlassian’s employees work in software engineering and design, accounting for over 50% of its 13,813 full-time workforce in June 2025.
Seems to me like this detail is particularly important though:
> [Atlassian] is not profitable and has recorded millions in losses every year since 2017, including a net loss of US$42m in the last three months of 2025, up from US$38m the prior year.
If a company isn't profitable without using AI, it's highly likely the execs don't know what their customers want and AI won't make the company profitable.
In the 90s and oughts everything was insecure by default. It was the golden age of teenage hackers. AI is ushering in a new golden age of insecure by default software.
AI reduces the need for expensive labor, payroll taxes and health insurance premiums, yet the cost of the company's service keeps going up. How does that happen?
Should've sent those 1600 people to fix their horrible performance of cloud apps, oh well I guess opening a jira ticket will now take not 5 but 10 seconds.
I'm retired now, but if I were looking for a job I'd try to find a company not using Atlassian products. In theory you're not supposed to use them as a (micro-) management tool, but companies like to do just that.
I’ve written on this. Jira is a code smell. The only people I’ve ever known who liked it were the people I don’t want to deal with. It’s a dream tool for someone who wants to make a career out of looking busy and inventing process, and a nightmare to everyone else. Its presence in an org tells me, in italic capitals, that this is going to hurt.
> "Our approach is not 'AI replaces people.' But it would be disingenuous to pretend AI doesn't change … the number of roles required in certain areas. It does," CEO Mike Cannon-Brookes said in a memo to employees.
“AI doesn’t replace people, but it’s silly to pretend it doesn’t replace people”, says man planning to use AI to replace people.
Well... companies who are laying off people on account of AI are telling markets that their chief value is NOT "innovation", it has been drudgery automation.
A company that prizes actually making customer lives better (i.e. "innovation") will salivate at the prospect of having ten of ten people, each using the Drudgery Automator, being creative using all that drugery-free time to actually touch grass with their customers and ship overnight value (i.e. actually, progressively, make things better for customers) and then improve it at a pace that blows minds (i.e. actually happy customers).
An company who's business is in fact about drudgery automation will kneecap itself by letting go nine of the ten.
When I see news like this, I feel very uneasy. It's not about a decrease in employment, but about the decrease in diversity in society. Compared to human diversity, I feel that the diversity of AI is very limited. I worry that everything in society will become uniform.
I do not see any reason to not believe Vince Gilligan saying coming up with Pluribus pre-AI and that it is not about AI, but it was impossible to watch that show without thinking about what a great metaphor it is for living surrounded by humans using AI to think and write for them.
> I worry that everything in society will become uniform
That’s why “slop” is so appropriate. It’s the runoff from the taps poured into a glass. Tastes like everything, nothing, definitely not pleasing, probably cheap, still effective at getting you inebriated.
Welp. Fuck Atlassian. Their resistance towards arbitrary RTO mandates and a people-first culture had me tolerate the weirdness of JIRA and Confluence, but now?
Fuck ‘em. Rolling my own using shelfware, kthxbai.
Seriously, you need a heck of a lot more than a random HN reply to give you Jira alternatives if you've been embedded into its ecosystem for any length of time - and my condolences if you have.
It's fine, just not stellar. It was terrible (UX, speed, consistency) ten years ago. It's better now - mostly gets out of people's way and just works. It doesn't delight me.
We've switched to Jetbrains Youtrack, it doesn't have as many features, but turns out nobody was using most of them anyway.
It's Jira + Confluence bundled together including SSO.
We (~200 devs) migrated from Jira to Youtrack 10 years ago, and its functionality has been more than enough. Honestly, I don't remember anyone ever seriously complaining about it, aside from maybe a few nitpicks. A very solid product.
So, like a lot of the tech industry this is simply a case of overhiring, overspending and general mismanagement. And like every other layoff announcement “AI” is a convenient scapegoat to hide executive dysfunction.
reply