Too expensive to fail?
Posted by karesx@reddit | ExperiencedDevs | View on Reddit | 88 comments
Have you ever participated in a project that was deemed too expensive to fail? Where the project is not ready yet but it is clear that it will never be profit positive. Still, your company (and the client) is pushing forward with it, because they have invested in so much that they just cannot afford to cancel it?
Comprehensive-Pea812@reddit
pretty much most banking migration projects
Admirable-Two-3143@reddit
Well i think these company owners need to read the book 'art of thinking clearly' then.
One-Marsupial2916@reddit
See sunk cost theory.
It’s one of the first things you learn in business school.
grilledcheex@reddit
I reflected the other day, and in 10 years as a dev I’m not sure I’ve ever worked on something that was profitable, at least not at the time. Scary thoughts. Is all my salary someone else’s loss?
liquidpele@reddit
It goes both ways, I've built multi-million dollar concepts for companies only to watch he whole company implode and get sold off because the people at the top didn't care and were just taking paychecks waiting to retire.
grilledcheex@reddit
I’ve probably saved my current employer multiples of my salary in avoided bad decisions so in that way my marginal contribution has been net positive.
Mike312@reddit
Tons of stuff I've worked on has saved the company money through automation or improving workflows. One department I built an interface for was supposed to grow from 3 to 5 people, and over 7 years they actually downsized to 2 because I automated so much of their work. The only reason it was 2 was to cover lunches, vacations, leaving to pick up kids from school, etc. 1 person could run the department in a pinch.
What does 3 people making $50k over 7 years cost a business? And that was just a 3-4 month project for me.
CornerDesigner8331@reddit
If a company that never turns a profit gets bought at a unicorn valuation, and the new parent company continues growing on paper, did anyone ever incur a loss?
cocacola999@reddit
Was at a company where mostly everything was a loss. Project I was on ended up turning a profit and they canned it... Ha
noblenomadas@reddit
Yep. That's called Sunk Cost Fallacy when people would rather put more resourses into a flop just because they've already invested versus pulling the plug which would be more beneficial.
tmac_arh@reddit
All day everyday. Then, it gets better. Because that "core" system was TETF (Too Expensive To Fail), the company goes out and either hires contractors, or buys multi-million dollar "low-code" integration tools to "fill in the gaps" of where the core software is drastically hurting the business. Next, because it still doesn't work, a whole entire team of "Snowflake Reporting Engineers" are hired to then do the same integration AGAIN, but this time bringing it into "Reporting Dashboards". Ce la vie.
Revision2000@reddit
Sure, see most government projects. These tend to go over time and over budget, but they usually simply must be delivered as there’s no other way forward.
Large enterprises can also suffer from the sunk cost fallacy.
Adept-Log3535@reddit
Cue the traditional defense industry. My ex-employer has an ongoing project that is already at 81%, or $80B, over budget. I am so happy that I got out
lordtosti@reddit
this is practically all government projects in all countries.
There is no accountability.
The person responsible for the project doesn’t want it to fail to save his own face, so he’ll keep spending money on it.
All the companies raking in all the money for sure don’t want it to stop. Usually the rates are also really high.
clusty1@reddit
This implies that gov projects HAVE to be profitable. There are other KPI than profit, like number of prevent gun deaths, number of tons of co2 taken form atmosphere.
This is not to say that gov projects are necessarily well managed….
thekwoka@reddit
True, and a lot of the really crazy overbudget ones are also projects that are quite cutting edge, to where it's hard to say to what degree mismanagement vs optimistic estimates impacts the additional cost.
lordtosti@reddit
cutting edge like?
thekwoka@reddit
F-35 and F-22 were pretty up there.
lordtosti@reddit
ok that is extremely niche compared to the thousands of other projects.
99% are of are probably just another web service and database as any other project.
yes, that can still be complex - but not hundreds of millions complex and for sure not billions of complex
thekwoka@reddit
I said the BIG ONEs.
The massive overruns
lordtosti@reddit
There is no “profit” for a government project as a concept, so you can’t anyway measure it.
There are a lot of projects that are hugely mismanaged. Everyone that have experience in the field knows these projects are hugely bureaucratic and for sure very hard mismanaged.
The government is just as inefficient as any other monopoly.
Spillz-2011@reddit
I don’t know about other countries but in the us they often have to go with lowest bidder. My mom runs a science team and they’ve had to contract to people they know can’t complete the work because they bid lower.
MissinqLink@reddit
Keeps spending money but continues to get the lowest bidding contractor
pfc-anon@reddit
Why not stay in and rake in at least a per cent.
AdvisedWang@reddit
Two reasons:
Its miserable
It looks terrible on a resume
nfigo@reddit
You have to work no matter where you go. Might as well work on a project that isn't trash. It's not like you will get paid more because your boss landed a bigger contract.
TopSwagCode@reddit
Yup this. Its not your money. You can give the best recommendations, but if they dont listen it isn't on you. I have seen a project going through 7 years without a single release.
Tacos314@reddit
That's how I view it, but really wish I got paid hourly to fix those mistakes
CornerDesigner8331@reddit
$80 billion is a rounding error on the F-35 overruns…
Politex99@reddit
I have not participated but I saw one. A game called Concord. $400 million. It was deemed too big to fail and it will the "Star Wars of gaming". It peaked < 700 online players when it launched on Steam.
SlightAddress@reddit
Sunken cost fallacy had been at least 80% of my work 😆 🤣
SlightAddress@reddit
The amount of time I've said "are you sure about this? I'm getting paid so it's no problem for me but are you sure this is a good idea?"
Southern-Reveal5111@reddit
Yes, I’ve seen it.
It was obvious the decision was bad, but they distorted the facts at every level. By the time it reached upper management, the truth was unrecognizable. It’s just another project meant to replace a legacy application. It was an eye-opener for me. Apparently, as long as bad news never gets delivered, everyone feels safe. Many will misjudge the messenger and will likely kill him. Most people just forward what they are told.
Live_Fall3452@reddit
My experience is that for every layer of management, the bad news gets very slightly sugarcoated. No one layer of sugar is really that big a deal, but by the time the news reaches the actual decision makers it’s almost pure sugar and the original news is close to gone.
Grubsnik@reddit
https://brucefwebster.com/2008/04/15/the-wetware-crisis-the-themocline-of-truth/
Just adding in this one for reference
CaptainKuzunoha@reddit
Very interesting read thank you.
liquidpele@reddit
Eh, I've seen that too, but it's really just a sign of nepotism and terrible management at a company. I've also been at good places where calling out even potential issues was the expectation. The people at the top set the terms for how everyone below them will conduct themselves.
dogo_fren@reddit
Wow, sounds like the Soviet Union.
light-triad@reddit
It's human nature once an organization gets big enough. The difference is that in a capitalist system. The company is either profitable in some other way, in which case it's not such a big deal they waste money, or the company goes bankrupt, which gets rid of the inefficiency.
In the Soviet Union, the rot grew for decades because the government didn't want to cancel projects that they deemed politically important. So it just kept getting worse and worse.
CornerDesigner8331@reddit
1000%. The parallels are disturbing if you know anything about their era of stagnation and gerontocracy.
I just hope the inevitable Balkanization is as peaceful. Although I doubt it will be.
cocacola999@reddit
I've currently living it. Corp merger makes it even more awkward as new company took over the project and moved loads of goalposts, sacked some key staff, etc. They've been telling exec it's all going great. It hasn't I think thee is an up coming set of deadlines we will miss and curious on how they will suger coat that. I'm betttig it will just be a toxic finger pointing at that point
Tacos314@reddit
Do you have skip level meetings? One reason these meetings exist is to prevent things like this.
cocacola999@reddit
Unfortunately for this context it is caused by the new org. They have replaced all management with their own.
CornerDesigner8331@reddit
I got killed a few times before I stopped trying to be the messenger. Unless people’s lives depend on the system in question, anyway.
PoopsCodeAllTheTime@reddit
I tend to speak up about problems because I believe that raising the issue is the only way to fix it in order to create a better produc..... Aaaand I'm fired again.
Altruistic_Tank3068@reddit
That's like all the government related project I used to work on.
It was not even too expensive to fail, it just didn't make any sense at first, but it was part of the current politics trends. As you can imagine those projects all crashed.
SeriousDabbler@reddit
I was on a project that was assessed as technically infeasible at the price point and reliability level that the customers were willing to pay for it, but the business had decided it was strategic and so we continued to flog it. Failing isn't the worst thing that can happen to a project
nfigo@reddit
Yeah, I've been in it.
Customers wanted refunds. So, we brought on more people to chase down issues.
None of this was necessary. The black hole is the team lead, who was sabotaging the project from the beginning. He was a darling of upper management and got shielded from consequences.
light-triad@reddit
Why was he sabotaging the project?
nfigo@reddit
I don't know for sure, but I think it was insecurity. He seemed to enjoy chaos and complicating things. He had to be the one to come up with an idea, or else he would find a way to fight it, especially if it came from outside the team. He talked forever in meetings and retaliated if he was interrupted. It seemed like a game to pretend that he was smarter than everyone else, like he was afraid of being found out.
North_Coffee3998@reddit
I've seen this too many times. They know that they are waaay above their head but their ego can't handle the truth. All those pointless meetings, the micromanagement, giving verbal instructions and only documenting what's convenient to them, adding unnecessary complexity, getting emotional for no reason, etc. It's all a tactic to mask their incompetence while stalling for time.
Dealing with people like that takes discipline. You have to be stoic and document everything. All of your actions have to be clearly traceable to their written instructions. When things don't go as planned calmly bring up the evidence while giving them outs (to confuse them and hide your intentions). The trick is to be defensive without them being certain of what you're doing, hence why being stoic is very important. Weaponize bureocracy against them. Looks hard at first until you get the hang of it.
If they can't get you fired, eventually they'll get tired and avoid you out of fear. After all, every interaction with you risks exposing them for the frauds that they are. Just keep being polite, stoic, and "helpful".
nfigo@reddit
That makes sense. I read this bit on the staffeng website, recently. I think if I could go back, I would have followed its advice. https://staffeng.com/guides/learn-to-never-be-wrong/
cocacola999@reddit
Oh I've worked with one of these types before. The whole set of leads were trying present evidence against the and our director still took many Many months to fire him. Damage was done though
TornadoFS@reddit
I didn't work on it but a project at the company I used to work at had so many SLA failures after delivery that our company had to pay the client more in fines than it received in support for the first few years.
And while that project was in development around 2 people quit per month (out of a team of 30 or so).
It was a project for the national postal office of the country, I think after 3 or so years the project was kind-of-okay and support for it was profitable. I don't know if the company ever made money out of it, but the amount of people capital lost set the company back a lot.
Mike312@reddit
So, there was this predication algorithm at the last place I worked at, can't really go into specifics. Basically, for 4 years it's been a running joke inside the team that it'll never launch. The kid who started writing it vibe-coded it before disappearing to a Christian camp in the woods for a year. It's caused one programmer to quit, another is on the verge of quitting. And the kicker is the client didn't even want the tool.
Here's more detail:
CEO spoke with one of our customers who stated they wanted a thing to do a certain mapping task.
The rest of the team was busy, so 17 YOE nepo hire develops the initial code. After a year of the kid coming in at 1pm every day and working on this algorithm entirely through ChatGPT and occasionally a little help from the rest of us, it worked! Sort of; like, all the variables are hard-coded, and the prediction moves at 400mph instead of 1-3mph, but still, progress!
Kid ships it to the main team a few weeks before he graduates from high school and goes off for a year to a weird backwoods Christian camp.
Rest of the team starts looking at the code, and it's exactly the nightmare vibe coded nonsense you think it is. CEO assigns one guy to work on it, and after about 6 months in he gets it actually functioning. Another guy gets brought onboard, and then another one. So like, 2 devs on month 7, 3 devs on month 9.
At 12 months the CEOs kid gets back from his weird camp and re-joins the project, but because all he's ever used is ChatGPT and doesn't know how to code (and it's been a year since he last used a computer, because weird Christian camp), he's dead weight. At 14 months they're cleaning it all up, and at 16 months they go to deliver it to the client...
...who says they already have a tool that does that, they were looking for something else.
At no point in the intervening 26 months did our CEO reach out to gather requirements or verify anything about this system or what it was supposed to do, or if the client even wanted it. At this point about 3,000 hours have been invested into this project that the CEO made up based on a 5-minute conversation, so he does what any incompetent CEO would do and doubles down.
I had left shortly before this happened, my team was the one he pulled the 3 programmers from, so I was burning out from keeping up with my work + their planned work.
I don't know exactly what they pivoted to; one of the guys told me, but it sounded dumb and the rest of the team was similarly as skeptical. It's been about 9 months since the pivot (and another 2,000 hours), and the guy that was initially brought on when the kid left for camp just quit. I hired him in 2022 and he had four decent-sized deliverables before he got transferred to this nightmare where he's sat for the last 3 years burning out. The second guy who got brought on is in a similar situation and the only reason he hasn't quit is the current job market.
RegrettableBiscuit@reddit
Sunk cost fallacy. Very common in software engineering.
HoratioWobble@reddit
I've worked on projects that weren't allowed to fail due to contractual and government ties.
That were going well, and then after a few years of working on the project arbitrarily outsourced the entire team or bought an external package and started again.
Oakw00dy@reddit
The cycle I've seen is: Start a project with vague requirements. Use up the budget. Deadmarch and failure to deliver. Everyone half competent quits. Move the goal posts. Fire the remaining ICs, promote the managers. Hire the cheapest poor unsuspecting sods to "finish the project". Set unrealistic completion deadlines. Rinse and repeat.
kopi32@reddit
I’ve worked in departments and thus projects that were never profitable in a large multinational corporation.
Instead of just narrowing in the facets of the project that make money and reducing down the team to a manageable amount, they double down: replace management, replace contractors, move operations to Europe from the US, add features that have no basis in user testing or marketing research, which is also a sham in a lot of ways (you find what you want to find). On top of all that, they move the goal posts all the time: we’re going to introduce this feature in x months, nope we’re going to focus on this instead. Best of all, we’re going to be profitable in x years. Too far down the line to have any merit, but just close enough that management has something they can get support for.
It’s just a never ending cycle of bad decisions and no accountability.
VRT303@reddit
Sunken cost fallacy
Flashy-Whereas-3234@reddit
While working for an agency I watched someone pour their mortgage into a project.
Our project managers and their CTO were both incompetent, but very much in cahoots lying about the quality of what we were delivering.
The agency touted the "agile approach" so strongly that each week of development had to have a customer demo, but internally we were spending so long making each demo that it was a hollow system with no real functionality. Everything was rushed because the PM couldn't nail down specs or designs, most of it needed replacing almost instantly.
After multiple late nights to deliver yet another godforsaken facade of a demo I decided to resign.
When they eventually ran out of runway for their big plane full of bullshit, the customer would go on to sue both the agency and their own CTO.
throwaway_0x90@reddit
What you're describing is:
Every experienced dev has seen this at least once.
dkHD7@reddit
What you describe is AI.
Far_Archer_4234@reddit
That phenominon is so common that has a term: sunk cost fallacy.
jl2352@reddit
Not to the degree of some of the stuff here. I once got sucked into a death march project to … change some frontend filters.
Users had a drop down listing regions of the world, and product wanted a slightly different list.
Went from a few months to multiple teams working well over a year to get it out. On release the platform would go down randomly everyday for four weeks. Three people quit over their frustrations from the project. One lead thought they were getting half way through. Only 60% was actually delivered.
As it was just ’changing some filters’, management just never gave up. It had to be done. Product later said if they knew it’d be half as painful they wouldn’t have pushed for it.
PracticallyPerfcet@reddit
Oh yea, the good ol’ sunk cost fallacy at play. I’ve seen it a handful of times over the years.
anuj_pardeshi_19@reddit
I’ve been there—sinking millions into a project we knew was a money pit, yet we kept going, blinded by sunk costs. In 2025, with AI hype and market pressure, I learned to pivot early—cut losses, refocus on what delivers value. Admitting failure isn’t defeat; it’s strategic clarity. DM me for my 7-day stress reset guide! Have you ever pushed through a doomed project? What made you stay?
Tango1777@reddit
Do you really care? You are not the founder, it's not your responsibility to deliver a product that generates profits. I don't really see any difference working for a project that'll fail or succeed. I do my job, take my money and switch the job if needed. Couldn't care less if a product is gonna succeed or not. I am not the company owner, not my problem. Obviously it's a high risk, high profit situation for the owners, sometimes you win, sometimes you lose. There is so much work in this business that half the companies could go bankrupt and I'd still have more work than available time.
JuiceChance@reddit
Yeap, it is common. Sometimes, it is the people that promised to build something, then they realized that they could not build it so they keep building a ridiciluous thing and keep telling management how amazing it is. Sometimes, it is the business that made stupid decision and carries on with it. Both situations happen quite often.
magpie882@reddit
Parent company refused to let a new subsidiary fail. Too much publicity and too many strings had been pulled.
After almost six years and churning through a ridiculous number of employees at the subsidiary and those dragged in from the main company, it is maybe going to finally go from negative EBITDA to positive.
Groove-Theory@reddit
yea this is called a Death March. A lot of sunk cost fallacies driven by big corporate and managerial 4heads who don't want to look stoopid and admit their own dumbassery.
Therefore, the dumbassery perpetuates itself.
Linaran@reddit
Did we accidentally describe the LLM hype? 😂
ConsiderationOk8231@reddit
Yes. Two words. AI.
birthnight@reddit
That's two letters.
OkSeaworthiness2727@reddit
The two words were before the letters. It was a trick statement.
DeterminedQuokka@reddit
Also most big data projects
positivcheg@reddit
Big data is a father of AI. I remember when 10 years ago “big data” was being laughed on the same way people laugh right now on any mention of AI (when AI is used as a synonym to LLMs).
Dro-Darsha@reddit
Or back when AI was synonymous with "basic statistics"
Tacos314@reddit
What sucks is you still keep working on those projects.
PoopsCodeAllTheTime@reddit
At this point it is difficult to find a project that isn't failing. All these garbage fires.... Last role I held was for a "unicorn startup".... They have been around for over ten years, haven't IPOd, recently asked for their TENTH ROUND OF INVESTMENT, and when I got hired.... I was working on their..... User enrollment?? What? They haven't figured out enrollment after 10 years? It was so broken too and most of the work involved shoving ads for adjacent products into the user's face. Disgusting.
Sparaucchio@reddit
I am in one of these projects and I am so thankful that investors are stuck in sunk cost fallacy, or we'd be jobless in a fricking rough market
Careful_Ad_9077@reddit
Yes.
The project failed, the company went bankrupt.
TolMera@reddit
If there is a scenario like this, it’s a notifyable scenario where the board should be notified and the shareholders informed. It should follow due diligence and such.
And an absolute defense in bringing it to the board, is being a shareholder of even a single share if it’s a publicly traded company.
Too big to fail, generally means someone somewhere high up the chain has their neck out too far. Time to cut it off.
PoopsCodeAllTheTime@reddit
Many many many agencies rely on the sunk cost fallacy to stay afloat for years.
arthoer@reddit
All of the projects that I worked on where the decision to build it came from people who wanted to hop on some kind of bandwagon has failed. eCommerce, blockchain, LLM. However the one that ALWAYS fails and ALWAYS comes back each year: WHITELABEL. Every director, project owner or board member has this grandiose idea to whitelabel a project when they come into position. Same goes for every junior, but they are ignored.
Fresh-String6226@reddit
Yes. People quit, leadership gets replaced, and it gets cancelled way too late. The worst case is when the leadership with their name attached to Big Project hide the problems and force it to ship anyway though.
justUseAnSvm@reddit
Yea, it feels like this on every project I've been on that has failed.
As a software engineer, we are the closest to the machine, we know what works, what doesn't work, and what has no possible way of working even in a million years. Therefore, our opinion of the project is going to lead what management thinks.
The moment we know things aren't working won't reach management until various KRs or indicators catch up with our observations. Sometimes that will take a month, but other times in large corporations it could be several quarters.
The way I think about it, is this is just a natural consequence of the doing engineering with any sort of scale, and being conservative about not cancelling anything too soon.
DeterminedQuokka@reddit
I thought I was and then they actually did eventually cancel it after some exceptionally bad pr several years later.