4 ms·

My Business Is Dying

89 points by 4pkjai 1mo ago

imhoguy 1mo ago
Content marketing is dying because content slop is massively generated now.
- 4pkjai 1mo ago
  I disagree, I think if you slop out your content it will be disregarded. It seems to be pretty easy to identify now with tools like Pangram.
  - imhoguy 1mo ago
    No, the problem is too much (sloppy) content dilutes good conent. Searching for good content became impossible, so instead people go straight to AI to ask.
    - Vivek-KY 1mo ago
      yes that's true
- dthedavid 1mo ago
  You can stand out by being authentic. Your face on the video, etc. Even writing it yourself genuinely instead of using an LLM. ex the article felt authentic.
  - imhoguy 1mo ago
    But how you get authentic and engaging for bank statement converter? People search for it to solve particular problem in the first place, not to hear your life story. And regarding face on video, I just got stuned with video of 106 years old giving life advices and all people but one in comment section didn't notice it is well generated.
    - dthedavid 1mo ago
      -> But how you get authentic and engaging for bank statement converter? People search for it to solve particular problem in the first place, not to hear your life story. That's the hard part :)
RMPR 1mo ago
> My friends tell me the job is basically managing AI agents and reading Pull Requests. That sounds like hell, and I’ll try to avoid it for as long as possible. It really is not that bad when you are in a company with a good culture. Leadership is gonna be AI piled everywhere, but imho that's how you learn where to use (and not to use) the technology.
- kevin_thibedeau 1mo ago
  Places that can't have code leave the work site still do not use AI.
  - Rebelgecko 1mo ago
    Some do, DOD is pushing grok
    - rootnod3 1mo ago
      Wonder how that happened...
      - glitchbot 1mo ago
        Someone bought a very expensive golf club membership?!
  - hobofan 1mo ago
    Most places I know that have strict on-premise/air-gap requirements are just investing in on-premise AI and open source harnesses instead. The only ones that don't are the typical places that are slow to adopt technology in general (usually because they can afford to due to historic moat).
    - sublinear 1mo ago
      Most places that must do things on-prem also have strict requirements and long deadlines. LLMs can be nice sometimes, but not necessary. Internal teams with "not invented here" syndrome currently working on that stuff are ultimately wasting money until the dust settles a bit more. Those long deadlines have nothing to do with implementation. It's the game of telephone and accountability bottlenecks. You still want to retain your most experienced devs and give them plenty of time to work. Writing the code was never the bottleneck. A good dev team has balanced experience of the business and the code, not just the code, and certainly not the kind of code an LLM generates. Most of the decisions by those devs happens outside the codebase.
losteric 1mo ago
I have a feeling search traffic is down and lots of people are diverting to using AI, or to one of the sloppy competitors. Conveying bank statements doesn’t seem like there’s much of a moat around the workflow.
wincy 1mo ago
I liked the article, although I have no idea what it even means to convert your bank statements. Is this a Chinese thing?
- Symbiote 1mo ago
  I assume it means converting PDFs transaction statements provided by banks into spreadsheets which accounting software can import.
  - navigate8310 1mo ago
    I'm surprised some banks don't provide a CSV export for statements, even if some do, they format it a XLSX file with images and meta descriptions in as header
  - qurren 1mo ago
    I just have Claude Code do this for me.
317070 1mo ago
Ironically, doesn't the "Get a job" paragraph have the solution? If you believe that is what the future of jobs at a company looks like, perhaps try to market your tool to agents? You could add an API which is perfect for agents to use. Something that survives the pull request review process.
- leftbehinds 1mo ago
  [dead]
fhd2 1mo ago
Hang in there. I've yet so see an even modestly successful business that didn't have this or some other bleak problem at one point in time. I'm sure they exist, I just haven't seen one yet :D My gut says this one can be turned around. Even if 90% of people just happily prefer ChatGPT for this (which I'd doubt), the remaining 10% is still a lot to fund a small venture very well. Just gotta find them effectively. Even the "giving up" route can be pretty juicy, there's companies that pay decent money for a "distressed asset". And there's also investment firms (like mine) that go for a minority stake and help turn things around.
altmanaltman 1mo ago
This is alright, but your main page is just trash. You mention you have 0 pages every 24 hours in the "Registered" free tier. You can't even check that but have time to write a post about how your business is dying? Like I understand you like content marketing but at least have your product ready for the potential customers that you might get via content marketing jfc
yevgenyhong 1mo ago
If it’s cheaper than having AI generate it directly, how about creating an MCP so AI can use this tool, and then trying connecting something like x402? It’s not exactly a subscription economy, but I think it could become a tool for AI
OsrsNeedsf2P 1mo ago
All I see is a company with a lot of users and good revenue. This looks like a great moat to build a business off of!
thoopring 1mo ago
[flagged]
einrealist 1mo ago
There is something I'd call Subscription Fatigue. People and businesses are asked to pay subscription fees left and right. If I convert my bank statements only each quarter in a spike or once or twice a year, why should I pay for each month or the entire year? Of course I don't know the real usage patterns of this application, but that's one scenario I'd think of.
- senectus1 1mo ago
  hell yes. I hit this wall about 2-3 years ago at least...
- johndubchak 1mo ago
  The perceptive utility of software declines when it is subscription-based in current times. A one-time fee for software is reasonable to a purchaser, renting/leasing software is no longer economical for the size of most phone apps...too little for too much. *Edit: Apparently I forgot how to spell.
- bayindirh 1mo ago
  Subscription fatigue is real, and I'm currently going through it, but it's not the existence of subscription-only model which nudges me to cancel a membership. If I pay for something, I want the tool to work for me. I can't use an application if it tries to steal my attention all the time. An upgrade pop-up there, a reminder notification there, and the application wears me down. I like quiet applications. It's there. I'll use it. An unobtrusive promo button for a new feature is OK, but constant nudging for the same thing is tiring. Dropbox used to do that with document signatures, they stopped now AFAICS. If I pay for your application, please be patient and respectful, because the moment I feel that your application is more about getting my money than providing me services, I'm out. The value I get is in the service you provide via your application. Not the existence of said application, so stop treating your users as cows in line waiting to be milked. This is another reason I'm self hosting as much as possible. I'm willing to put in the work to maintain an application rather than seeing constant banners and pop-ups wedged in to milk me more.
  - 1mo ago
JV00 1mo ago
I like this writing style, dry and direct. Refreshing after reading ai output all the time.
- FailMore 1mo ago
  Agreed
bombcar 1mo ago
AI kills SaaS dead - because most will use AI to do what this does directly (presumably convert bank statements), and the rest will use AI to write a non-AI tool that does it likely just as well.
- jbs789 1mo ago
  I did this… I have a specific system-generated PDF bank statement and it didn’t take too long to write a parsing program, which would have taken much longer before.
dannypostma 1mo ago
Seeing this with all my founder friends. Combination of Google AI summaries just sending much less traffic and AI tools being able to do small tasks by themself, rending a ton of SaaS less used.
jdw64 1mo ago
Taking my homepage as an example, when I allowed AI bots, I got about 1,200 visitors a day. But when I blocked them, that number dropped to around 140. It seems like over 70% of internet traffic these days is dominated by AI bots.
jbs789 1mo ago
> When a user cancels a subscription Stripe asks them for a reason. At the moment I don’t really do anything with the cancellation reason, actually this is the first time I’ve ever looked at the data. This comes across as very unserious. If you’re truly trying to build a business then this seems like one of the best signals/datapoints you can get…
- AdamN 1mo ago
  I often think of my experience at Citibank 15 years ago when I closed my account at the branch. The teller said 'sure thing, just put your debit card into the device and enter your pin'. I did that and he said ok I'll close the account and give you cash? I said 'yes' and then it was done. He said, "have a great day!" It was one of the best customer experiences of my life to not have to answer a series of 'why are you leaving?' questions and I would like to head back to Citibank just because of it. Unfortunately I don't live in NYC anymore so can't really do that.
  - slazaro 1mo ago
    I agree that most companies are awful at trying to get you to stay when you tell them to leave, but I'd be disappointed if a company didn't care at all to know why I left. If I'm leaving I already have a bad feeling about them, and I would have a worse image of them if they didn't care why. Ideally they wouldn't pressure at all for me to stay and cancellation would be a quick process, and I would be happy to tell them why I left. Either to vent, or to at least get the (most likely misguided) feeling that they might do something about it. But if there's no asking as part of their process, I'd definitely think worse of them.
    - jbs789 1mo ago
      Yeah I like to make the question optional. No impact on their exit unless they want to share something.
MagicMoonlight 1mo ago
[dead]
hnthrow10282910 1mo ago
This was a really honest and nice read. I also appreciated the last paragraph. Personally I work at a FAANG adjacent big tech company and dream of quitting every fucking second
nostromo 1mo ago
A lot of these super-simple utility apps will die off with AI. Two years ago I paid a little app to extract the member list of a WhatsApp community I manage. Now I just ask Codex to do it. It sucks for the founders no doubt. But it’s better for the consumer not to have a million paid utility apps for solving small nuisances.
unmole 1mo ago
> People are using AI to convert their bank statements. This works pretty well if you only have a few pages of bank statements to convert. If you have hundreds of pages, then you’ll need to pay for AI to do the conversion. Or you could have AI write a script to handle the conversion for your specific bank. It'll be faster than any service and free forever.
- kodoman 1mo ago
  This is often what I find people do with AI now, give it the task to do when getting it to writing a script is the correct solution. Further the issue of exposing all your sensitive data to the AI provider.
cjs_ac 1mo ago
The author’s product extracts data from bank statement PDFs and puts it in XLSX format. Both of the banks I use provide CSV data export from their apps. I’m not surprised the business is struggling - the author hasn’t found a deep enough problem to solve.
- dchuk 1mo ago
  Not to mention all of the SOTA AI subscriptions handle this perfectly now when there aren’t csv exports
bob1029 1mo ago
> Last time I worked at a company was November 2020. I love programming, but I really do not enjoy programming at a company. Things were bad in 2020, but I feel like they’re worse now. They're much worse now, but that also creates more opportunity than ever before. Chaos truly can be a ladder if you take advantage of it and endure the unhappy parts. Having a bit of savings to work with can make all the difference. You can focus on emerging opportunities without melting down over your next paycheck. The concept of W-2 employment is completely dead to me now. Especially for this kind of technology work. I charge per outcome now (1099). I am no longer interested in being provided what amounts to hazard pay for the job of interacting with an adult day care center every day. Walking in as a hired gun with no political agenda and a clear mission feels great compared to where I was before. When I do participate in the clients' meetings, I generally get away with murder because I am laser focused on the outcome itself (my compensation) rather than belaboring whatever technical details behind it. If I'm on a call these days, there's an explicit reason for it. The business can sense that I am aligned with their goals, so the proverbial seas tend to be parted whenever I unmute my microphone. The clients' salaried developers loathe me to the extent that it's creating even more opportunities. The amount of unqualified ego tripping in tech has become astounding and can serve as a powerful ally for humble wizards.
- rightbyte 1mo ago
  It is not ego tripping. They probably loathe you because you turn the power structures even more in capital's favour by undercutting or overriding them.
  - IsTom 1mo ago
    Or leaving shitty code for others to fix later, like a lot of contractors do.
    - tonyarkles 1mo ago
      I’m smiling a bit reading this thread because I went essentially the opposite way: started contracting in 2010 when my grad school stipend ran out, in Feb 2020 (kind of lucky timing…) one of my clients convinced me to come on as FTE #2 at their startup (I’d done a 6-month prototyping gig for them in 2019 that was responsible for them unlocking their first tranche of funding). Have stuck around since then because the work is compelling and to some degree I’ve shaped the engineering org into what I want it to be, even though we’ve hired people above me (it’s kind of neat and weird hiring your own manager so that you no longer report direct to the CEO) Anyway, as far as I can tell, the engineering teams that I was paid to help, I believe, generally loved me. The work I’d do generally fell into one of three buckets: - early-stage prototyping work, where there was no engineering team yet and I was laying the foundation for v1 of a product - modernization work, where a company has a product line that’s sitting generating revenue but there’s no engineering capacity to actually make v2. This was often more on the embedded side but occasionally stuff like pre-ARC Objective C -> Swift. Everyone’s too busy working on some other thing and the good-but-crusty product is starting to die due to lack of attention - rescue work, where the engineering organization (management + ICs) have gotten themselves into a situation they’ve realized they can’t get themselves out of on their own. Often performance, scaling, or velocity-related, almost always with an element of pathological engineering culture. The first two cases aren’t as relevant, but the third type was always a hearts & minds campaign. The outcome was two-fold: fix the software, fix the team. First step: take the team out for lunch without their manager. To the loathing comment above, not everyone would take me up on the offer, but I could usually get enough of a critical mass together. Two objectives: convince them that I do not, under any circumstances, want to work with them long-term; convince them that I was on their side. Get them talking about what’s wrong, ostensibly about the product, but this would frequently pivot to the structural organizational problems that led to the product being how it was. From there, put together two plans. Plan 1 was for how to tackle the software problem. Plan 2 was how to fix the engineering org. Typical situation was they had some ideas on the first part and LOTS to say about the second part. The connection between the two was they often had a good high-level idea about how to fix the software but didn’t have the specific skill to do the fix and didn’t have time to learn. Then I’d start working both problems in parallel: knock out a quick win on the software side, do a lunch & learn with the team and their manager to explain what I’ve done, what tools I used if I used anything exotic, discuss next steps, and discuss whether someone else on the team wants to work together on the next thing. Meanwhile, I’m watching the team (incl. management) and the complaints that I got from them over lunch and while working with them. Looking to see which of those complaints I can see in the day-to-day work environment and who to… blame isn’t the right word, but figuring out who I need to nudge in which direction to make the problems that led to my involvement go away. This usually went through a few iteration loops but I was always clear to everyone involved that my goal was for the team to not need me anymore. That was the outcome: problem solved, team can solve the same kinds of problems on their own without me. I’m available if you ever need to hire me again, I’m here if you get stuck on something tactical and can help out for a few days here and there. If my goal was to get fired, how do I know that I did good? Direct feedback from ICs was one way. I smoked a fair bit back then and the number of times I had people come up to me outside during a coffee/smoke break and just say “hey man, I really appreciate what you’re doing here, I’ve learned a lot about ____”. The other one: referrals. Maybe 20% of the work was inbound LinkedIn cold calls, while the other 80% was referrals from existing clients (or callbacks) with about a 50-50 mix of that being management or IC referrals. Work with a team for 3 months, a couple months later get an email “hey so I was out for beer with XXX and I think you might be able to help me…”
apt-apt-apt-apt 1mo ago
I'll be launching in a week. "Claude, a business is making $30K/mo converting bank statements! Use all of my max 20x plan to build a BETTER version. Must have animated per-page progress indicators"
librasteve 1mo ago
It’s fascinating to read all the tech feedback in the comments here, but I think this business is failing because it has yet to achieve true product market fit. Any half decent product, with a nice website, a real solution and some SEO can get in front of say 100,000 prospects these days. Here I imagine discovery is mainly from Googling “convert bank statements”, and conversion from prospects to client is straightforward for 1-2% of views. But a conversion niche like this is dynamic - you basically luck into it and unless you actively find and occupy niches faster than the competition discovers them and stomps on you, your “business” is on a time fuze. For example, if someone like Xero where this is a feature discovers the keyword combo and does some paid ads. Or if Google pushes AI results up and your page goes below the fold. My expectation is that this latter situation is happening everywhere as Google AI search starts to strangle everything. Leverage AI to be your CMO. It should be telling you things like (i) your product is likely perceived as too technical to break into the general market, (ii) you need a much better handle in target customer profile (demographic, messages), (iii) you need a customer relevant sustainable USP and (iv) you need a distinctive and memorable brand. The ultimate win is to create market pull from customer to customer recommendations.
- alexaholic 1mo ago
  > Any half decent product, with a nice website, a real solution and some SEO can get in front of say 100,000 prospects these days A lot of the Internet says otherwise — especially these days. Care to share your sauce?
  - xnx 1mo ago
    Easy to get in front of that many people. Hard to do so profitably. If Google is doing its job, it should be taking most of the margin.
tchalla 1mo ago
> Sharing revenue data is a good way to get attention on Twitter, but for this business the Twitter attention doesn’t really translate into sales. I stopped sharing revenue data because a lot of people started using AI to build their own bank statement converter applications. A lot of build in public is theater with MRR porn which has no basis in ground reality creating noise. Now the agentic world is coming after it. Build in public has turned into display in public these days.
ed_mercer 1mo ago
OP was doing 30K USD MRR, not sure why he needs to find a job at all. Did he spend it all?
Swankivo 1mo ago
Hundreds of SaaS companies disappeared in Korea too. They don't even make the news. The examples that do come up now and then, Wrtn, Liner, Qanda, those are the ones that survived, so the press writes them down. A friend's company took a huge valuation cut and ran a round on very bad terms. In the end an existing investor re-upped one last time and that was the lifeline. It's not that the strong survive, it's a market where whoever survives is strong.
ernsheong 1mo ago
Basically AI can do the job now. This kind of tool has been nullified by Gemini Notebook, Claude Cowork, etc. or any LLM chat for that matter
leftbehinds 1mo ago
[dead]
techdecoded 1mo ago
[dead]
teeray 1mo ago
> AI chat bots can convert PDF to Excel files, but they’re slower and more expensive. I have a feeling a lot of users are using free tier AI chatbots to do their conversions. I’ve considered this, but gosh does it make me nervous to potentially put that info into training sets. Instead I’ve left that task to local LLMs.
- dumbmrblah 1mo ago
  There is no moat. You can vibecode an app that does this conversion in 30 minutes and never need to upload it to a AI again. I'm not saying this right or wrong, but it is what a tech savy (e.g HN audience) has and will be doing. ' Perhaps we aren't the market for this webapp but I don't really the value it in post 2025
pan69 1mo ago
Who uploads their bank statements into some random website?

89 points by 4pkjai 1mo ago

imhoguy 1mo ago

Content marketing is dying because content slop is massively generated now.

4pkjai 1mo ago
I disagree, I think if you slop out your content it will be disregarded. It seems to be pretty easy to identify now with tools like Pangram.
- imhoguy 1mo ago
  No, the problem is too much (sloppy) content dilutes good conent. Searching for good content became impossible, so instead people go straight to AI to ask.
  - Vivek-KY 1mo ago
    yes that's true
dthedavid 1mo ago
You can stand out by being authentic. Your face on the video, etc. Even writing it yourself genuinely instead of using an LLM. ex the article felt authentic.
- imhoguy 1mo ago
  But how you get authentic and engaging for bank statement converter? People search for it to solve particular problem in the first place, not to hear your life story. And regarding face on video, I just got stuned with video of 106 years old giving life advices and all people but one in comment section didn't notice it is well generated.
  - dthedavid 1mo ago
    -> But how you get authentic and engaging for bank statement converter? People search for it to solve particular problem in the first place, not to hear your life story. That's the hard part :)

RMPR 1mo ago

> My friends tell me the job is basically managing AI agents and reading Pull Requests. That sounds like hell, and I’ll try to avoid it for as long as possible. It really is not that bad when you are in a company with a good culture. Leadership is gonna be AI piled everywhere, but imho that's how you learn where to use (and not to use) the technology.

kevin_thibedeau 1mo ago
Places that can't have code leave the work site still do not use AI.
- Rebelgecko 1mo ago
  Some do, DOD is pushing grok
  - rootnod3 1mo ago
    Wonder how that happened...
    - glitchbot 1mo ago
      Someone bought a very expensive golf club membership?!
- hobofan 1mo ago
  Most places I know that have strict on-premise/air-gap requirements are just investing in on-premise AI and open source harnesses instead. The only ones that don't are the typical places that are slow to adopt technology in general (usually because they can afford to due to historic moat).
  - sublinear 1mo ago
    Most places that must do things on-prem also have strict requirements and long deadlines. LLMs can be nice sometimes, but not necessary. Internal teams with "not invented here" syndrome currently working on that stuff are ultimately wasting money until the dust settles a bit more. Those long deadlines have nothing to do with implementation. It's the game of telephone and accountability bottlenecks. You still want to retain your most experienced devs and give them plenty of time to work. Writing the code was never the bottleneck. A good dev team has balanced experience of the business and the code, not just the code, and certainly not the kind of code an LLM generates. Most of the decisions by those devs happens outside the codebase.

losteric 1mo ago

I have a feeling search traffic is down and lots of people are diverting to using AI, or to one of the sloppy competitors. Conveying bank statements doesn’t seem like there’s much of a moat around the workflow.

wincy 1mo ago

I liked the article, although I have no idea what it even means to convert your bank statements. Is this a Chinese thing?

Symbiote 1mo ago
I assume it means converting PDFs transaction statements provided by banks into spreadsheets which accounting software can import.
- navigate8310 1mo ago
  I'm surprised some banks don't provide a CSV export for statements, even if some do, they format it a XLSX file with images and meta descriptions in as header
- qurren 1mo ago
  I just have Claude Code do this for me.

317070 1mo ago

Ironically, doesn't the "Get a job" paragraph have the solution? If you believe that is what the future of jobs at a company looks like, perhaps try to market your tool to agents? You could add an API which is perfect for agents to use. Something that survives the pull request review process.

leftbehinds 1mo ago
[dead]

fhd2 1mo ago

Hang in there. I've yet so see an even modestly successful business that didn't have this or some other bleak problem at one point in time. I'm sure they exist, I just haven't seen one yet :D My gut says this one can be turned around. Even if 90% of people just happily prefer ChatGPT for this (which I'd doubt), the remaining 10% is still a lot to fund a small venture very well. Just gotta find them effectively. Even the "giving up" route can be pretty juicy, there's companies that pay decent money for a "distressed asset". And there's also investment firms (like mine) that go for a minority stake and help turn things around.

altmanaltman 1mo ago

This is alright, but your main page is just trash. You mention you have 0 pages every 24 hours in the "Registered" free tier. You can't even check that but have time to write a post about how your business is dying? Like I understand you like content marketing but at least have your product ready for the potential customers that you might get via content marketing jfc

yevgenyhong 1mo ago

If it’s cheaper than having AI generate it directly, how about creating an MCP so AI can use this tool, and then trying connecting something like x402? It’s not exactly a subscription economy, but I think it could become a tool for AI

OsrsNeedsf2P 1mo ago

All I see is a company with a lot of users and good revenue. This looks like a great moat to build a business off of!

thoopring 1mo ago

[flagged]

einrealist 1mo ago

There is something I'd call Subscription Fatigue. People and businesses are asked to pay subscription fees left and right. If I convert my bank statements only each quarter in a spike or once or twice a year, why should I pay for each month or the entire year? Of course I don't know the real usage patterns of this application, but that's one scenario I'd think of.

senectus1 1mo ago
hell yes. I hit this wall about 2-3 years ago at least...
johndubchak 1mo ago
The perceptive utility of software declines when it is subscription-based in current times. A one-time fee for software is reasonable to a purchaser, renting/leasing software is no longer economical for the size of most phone apps...too little for too much. *Edit: Apparently I forgot how to spell.
bayindirh 1mo ago
Subscription fatigue is real, and I'm currently going through it, but it's not the existence of subscription-only model which nudges me to cancel a membership. If I pay for something, I want the tool to work for me. I can't use an application if it tries to steal my attention all the time. An upgrade pop-up there, a reminder notification there, and the application wears me down. I like quiet applications. It's there. I'll use it. An unobtrusive promo button for a new feature is OK, but constant nudging for the same thing is tiring. Dropbox used to do that with document signatures, they stopped now AFAICS. If I pay for your application, please be patient and respectful, because the moment I feel that your application is more about getting my money than providing me services, I'm out. The value I get is in the service you provide via your application. Not the existence of said application, so stop treating your users as cows in line waiting to be milked. This is another reason I'm self hosting as much as possible. I'm willing to put in the work to maintain an application rather than seeing constant banners and pop-ups wedged in to milk me more.
- 1mo ago

JV00 1mo ago

I like this writing style, dry and direct. Refreshing after reading ai output all the time.

FailMore 1mo ago
Agreed

bombcar 1mo ago

AI kills SaaS dead - because most will use AI to do what this does directly (presumably convert bank statements), and the rest will use AI to write a non-AI tool that does it likely just as well.

jbs789 1mo ago
I did this… I have a specific system-generated PDF bank statement and it didn’t take too long to write a parsing program, which would have taken much longer before.

dannypostma 1mo ago

Seeing this with all my founder friends. Combination of Google AI summaries just sending much less traffic and AI tools being able to do small tasks by themself, rending a ton of SaaS less used.

jdw64 1mo ago

Taking my homepage as an example, when I allowed AI bots, I got about 1,200 visitors a day. But when I blocked them, that number dropped to around 140. It seems like over 70% of internet traffic these days is dominated by AI bots.

jbs789 1mo ago

> When a user cancels a subscription Stripe asks them for a reason. At the moment I don’t really do anything with the cancellation reason, actually this is the first time I’ve ever looked at the data. This comes across as very unserious. If you’re truly trying to build a business then this seems like one of the best signals/datapoints you can get…

AdamN 1mo ago
I often think of my experience at Citibank 15 years ago when I closed my account at the branch. The teller said 'sure thing, just put your debit card into the device and enter your pin'. I did that and he said ok I'll close the account and give you cash? I said 'yes' and then it was done. He said, "have a great day!" It was one of the best customer experiences of my life to not have to answer a series of 'why are you leaving?' questions and I would like to head back to Citibank just because of it. Unfortunately I don't live in NYC anymore so can't really do that.
- slazaro 1mo ago
  I agree that most companies are awful at trying to get you to stay when you tell them to leave, but I'd be disappointed if a company didn't care at all to know why I left. If I'm leaving I already have a bad feeling about them, and I would have a worse image of them if they didn't care why. Ideally they wouldn't pressure at all for me to stay and cancellation would be a quick process, and I would be happy to tell them why I left. Either to vent, or to at least get the (most likely misguided) feeling that they might do something about it. But if there's no asking as part of their process, I'd definitely think worse of them.
  - jbs789 1mo ago
    Yeah I like to make the question optional. No impact on their exit unless they want to share something.

MagicMoonlight 1mo ago

[dead]

hnthrow10282910 1mo ago

This was a really honest and nice read. I also appreciated the last paragraph. Personally I work at a FAANG adjacent big tech company and dream of quitting every fucking second

nostromo 1mo ago

A lot of these super-simple utility apps will die off with AI. Two years ago I paid a little app to extract the member list of a WhatsApp community I manage. Now I just ask Codex to do it. It sucks for the founders no doubt. But it’s better for the consumer not to have a million paid utility apps for solving small nuisances.

unmole 1mo ago

> People are using AI to convert their bank statements. This works pretty well if you only have a few pages of bank statements to convert. If you have hundreds of pages, then you’ll need to pay for AI to do the conversion. Or you could have AI write a script to handle the conversion for your specific bank. It'll be faster than any service and free forever.

kodoman 1mo ago
This is often what I find people do with AI now, give it the task to do when getting it to writing a script is the correct solution. Further the issue of exposing all your sensitive data to the AI provider.

cjs_ac 1mo ago

The author’s product extracts data from bank statement PDFs and puts it in XLSX format. Both of the banks I use provide CSV data export from their apps. I’m not surprised the business is struggling - the author hasn’t found a deep enough problem to solve.

dchuk 1mo ago
Not to mention all of the SOTA AI subscriptions handle this perfectly now when there aren’t csv exports

bob1029 1mo ago

> Last time I worked at a company was November 2020. I love programming, but I really do not enjoy programming at a company. Things were bad in 2020, but I feel like they’re worse now. They're much worse now, but that also creates more opportunity than ever before. Chaos truly can be a ladder if you take advantage of it and endure the unhappy parts. Having a bit of savings to work with can make all the difference. You can focus on emerging opportunities without melting down over your next paycheck. The concept of W-2 employment is completely dead to me now. Especially for this kind of technology work. I charge per outcome now (1099). I am no longer interested in being provided what amounts to hazard pay for the job of interacting with an adult day care center every day. Walking in as a hired gun with no political agenda and a clear mission feels great compared to where I was before. When I do participate in the clients' meetings, I generally get away with murder because I am laser focused on the outcome itself (my compensation) rather than belaboring whatever technical details behind it. If I'm on a call these days, there's an explicit reason for it. The business can sense that I am aligned with their goals, so the proverbial seas tend to be parted whenever I unmute my microphone. The clients' salaried developers loathe me to the extent that it's creating even more opportunities. The amount of unqualified ego tripping in tech has become astounding and can serve as a powerful ally for humble wizards.

rightbyte 1mo ago
It is not ego tripping. They probably loathe you because you turn the power structures even more in capital's favour by undercutting or overriding them.
- IsTom 1mo ago
  Or leaving shitty code for others to fix later, like a lot of contractors do.
  - tonyarkles 1mo ago
    I’m smiling a bit reading this thread because I went essentially the opposite way: started contracting in 2010 when my grad school stipend ran out, in Feb 2020 (kind of lucky timing…) one of my clients convinced me to come on as FTE #2 at their startup (I’d done a 6-month prototyping gig for them in 2019 that was responsible for them unlocking their first tranche of funding). Have stuck around since then because the work is compelling and to some degree I’ve shaped the engineering org into what I want it to be, even though we’ve hired people above me (it’s kind of neat and weird hiring your own manager so that you no longer report direct to the CEO) Anyway, as far as I can tell, the engineering teams that I was paid to help, I believe, generally loved me. The work I’d do generally fell into one of three buckets: - early-stage prototyping work, where there was no engineering team yet and I was laying the foundation for v1 of a product - modernization work, where a company has a product line that’s sitting generating revenue but there’s no engineering capacity to actually make v2. This was often more on the embedded side but occasionally stuff like pre-ARC Objective C -> Swift. Everyone’s too busy working on some other thing and the good-but-crusty product is starting to die due to lack of attention - rescue work, where the engineering organization (management + ICs) have gotten themselves into a situation they’ve realized they can’t get themselves out of on their own. Often performance, scaling, or velocity-related, almost always with an element of pathological engineering culture. The first two cases aren’t as relevant, but the third type was always a hearts & minds campaign. The outcome was two-fold: fix the software, fix the team. First step: take the team out for lunch without their manager. To the loathing comment above, not everyone would take me up on the offer, but I could usually get enough of a critical mass together. Two objectives: convince them that I do not, under any circumstances, want to work with them long-term; convince them that I was on their side. Get them talking about what’s wrong, ostensibly about the product, but this would frequently pivot to the structural organizational problems that led to the product being how it was. From there, put together two plans. Plan 1 was for how to tackle the software problem. Plan 2 was how to fix the engineering org. Typical situation was they had some ideas on the first part and LOTS to say about the second part. The connection between the two was they often had a good high-level idea about how to fix the software but didn’t have the specific skill to do the fix and didn’t have time to learn. Then I’d start working both problems in parallel: knock out a quick win on the software side, do a lunch & learn with the team and their manager to explain what I’ve done, what tools I used if I used anything exotic, discuss next steps, and discuss whether someone else on the team wants to work together on the next thing. Meanwhile, I’m watching the team (incl. management) and the complaints that I got from them over lunch and while working with them. Looking to see which of those complaints I can see in the day-to-day work environment and who to… blame isn’t the right word, but figuring out who I need to nudge in which direction to make the problems that led to my involvement go away. This usually went through a few iteration loops but I was always clear to everyone involved that my goal was for the team to not need me anymore. That was the outcome: problem solved, team can solve the same kinds of problems on their own without me. I’m available if you ever need to hire me again, I’m here if you get stuck on something tactical and can help out for a few days here and there. If my goal was to get fired, how do I know that I did good? Direct feedback from ICs was one way. I smoked a fair bit back then and the number of times I had people come up to me outside during a coffee/smoke break and just say “hey man, I really appreciate what you’re doing here, I’ve learned a lot about ____”. The other one: referrals. Maybe 20% of the work was inbound LinkedIn cold calls, while the other 80% was referrals from existing clients (or callbacks) with about a 50-50 mix of that being management or IC referrals. Work with a team for 3 months, a couple months later get an email “hey so I was out for beer with XXX and I think you might be able to help me…”

apt-apt-apt-apt 1mo ago

I'll be launching in a week. "Claude, a business is making $30K/mo converting bank statements! Use all of my max 20x plan to build a BETTER version. Must have animated per-page progress indicators"

librasteve 1mo ago

It’s fascinating to read all the tech feedback in the comments here, but I think this business is failing because it has yet to achieve true product market fit. Any half decent product, with a nice website, a real solution and some SEO can get in front of say 100,000 prospects these days. Here I imagine discovery is mainly from Googling “convert bank statements”, and conversion from prospects to client is straightforward for 1-2% of views. But a conversion niche like this is dynamic - you basically luck into it and unless you actively find and occupy niches faster than the competition discovers them and stomps on you, your “business” is on a time fuze. For example, if someone like Xero where this is a feature discovers the keyword combo and does some paid ads. Or if Google pushes AI results up and your page goes below the fold. My expectation is that this latter situation is happening everywhere as Google AI search starts to strangle everything. Leverage AI to be your CMO. It should be telling you things like (i) your product is likely perceived as too technical to break into the general market, (ii) you need a much better handle in target customer profile (demographic, messages), (iii) you need a customer relevant sustainable USP and (iv) you need a distinctive and memorable brand. The ultimate win is to create market pull from customer to customer recommendations.

alexaholic 1mo ago
> Any half decent product, with a nice website, a real solution and some SEO can get in front of say 100,000 prospects these days A lot of the Internet says otherwise — especially these days. Care to share your sauce?
- xnx 1mo ago
  Easy to get in front of that many people. Hard to do so profitably. If Google is doing its job, it should be taking most of the margin.

tchalla 1mo ago

> Sharing revenue data is a good way to get attention on Twitter, but for this business the Twitter attention doesn’t really translate into sales. I stopped sharing revenue data because a lot of people started using AI to build their own bank statement converter applications. A lot of build in public is theater with MRR porn which has no basis in ground reality creating noise. Now the agentic world is coming after it. Build in public has turned into display in public these days.

ed_mercer 1mo ago

OP was doing 30K USD MRR, not sure why he needs to find a job at all. Did he spend it all?

Swankivo 1mo ago

Hundreds of SaaS companies disappeared in Korea too. They don't even make the news. The examples that do come up now and then, Wrtn, Liner, Qanda, those are the ones that survived, so the press writes them down. A friend's company took a huge valuation cut and ran a round on very bad terms. In the end an existing investor re-upped one last time and that was the lifeline. It's not that the strong survive, it's a market where whoever survives is strong.

ernsheong 1mo ago

Basically AI can do the job now. This kind of tool has been nullified by Gemini Notebook, Claude Cowork, etc. or any LLM chat for that matter

leftbehinds 1mo ago

[dead]

techdecoded 1mo ago

[dead]

teeray 1mo ago

> AI chat bots can convert PDF to Excel files, but they’re slower and more expensive. I have a feeling a lot of users are using free tier AI chatbots to do their conversions. I’ve considered this, but gosh does it make me nervous to potentially put that info into training sets. Instead I’ve left that task to local LLMs.

dumbmrblah 1mo ago
There is no moat. You can vibecode an app that does this conversion in 30 minutes and never need to upload it to a AI again. I'm not saying this right or wrong, but it is what a tech savy (e.g HN audience) has and will be doing. ' Perhaps we aren't the market for this webapp but I don't really the value it in post 2025

pan69 1mo ago

Who uploads their bank statements into some random website?