My Business Is Dying

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.

That's the hard part :)

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.

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.

Sure, the slow behemoths that you talk about exist as well, but I am seeing many companies that have successfully turned compliance into just another kind of process that isn't slowing down the product development lifecycle. And within those process steps they are also able to leverage AI to speed up e.g. review processes.

[0]: https://github.com/EratoLab/erato

I have doubts

This is the magical sci-fi future I always dreamed about, and it's getting better every quarter.

The amount of stuff I can do now, in terms of time and scope and volume, is unreal.

You could add an API which is perfect for agents to use. Something that survives the pull request review process.

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.

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.

I hit this wall about 2-3 years ago at least...

*Edit: Apparently I forgot how to spell.

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.

[1]: https://calmtech.com/

I don't believe such future will arrive by itself, but we can build a calm tech stack for ourselves. I successfully pruned much of the noisier stuff from my environment, and cutting down subscriptions left and right.

Unfortunately, some of the subscriptions are really useful, or I like how they present the thing I'm working on, so they stay for the time being.

Thanks for the link. I'll be diving into this.

But really that cacophony of upgrades etc isn’t about user control, but just companies reminding you they exist. It’s spam, and like spam fighting, agents will not be able to handle it adequately.

macOS and iOS doing updates when you are not looking is good because they store your app state most of the time.

Some macOS programs' "We have downloaded the update, it'll update silently when you exit" is also nice and kind way of doing it.

Firefox recently remedied its "We updated in the background but broke the runtime" problem for some months now on Linux and macOS, from which I observe.

However nudging, aggressive updates, forcing user to stop or take breaks are not calm technology. I accept that wholeheartedly.

People don't appreciate enough that releasing new software has a cost. Upgrading is risky and disruptive. Even on iOS, where app upgrades are as low effort as possible, they're a nuisance; I'm occasionally unable to open an app I need because it's mid-update. One time, I was using an app, closed it to get some information from another app, and when I wanted to go back to the first app I couldn't because it had started an auto update.

Software updates are a necessary evil. As such, they should be as infrequent as is reasonably possible.

If you want to go nuclear, you can disable update notifications altogether under administrative settings, I believe.

On the other hand, I mostly charge my iOS devices at night, and they handle updates & backups when they are charging. I didn't see an app updating midday for years, I believe.

Nextcloud is a bit too noisy on the app updates department, though, we can all agree on that.

Almost everything else in my life is a one time purchase. I pay for the thing, I use it forever. I would never buy a vacuum cleaner subscription. Or a pressure washer subscription. Why should I need to endlessly pay for software?

The usual excuses from developers:

1. But I need to maintain the server and that's a monthly cost!

Nope. Tethering your software to a server was your choice, not mine. I shouldn't have to pay for your decision. I have lots of perfectly good software on my computer that work fine without your server.

2. But it costs money to constantly be developing and publishing updates!

Nope. Nobody is asking you to constantly change your software. Generally, I don't want your updates. The software met my needs when I bought it, and I don't want you to change it, and I certainly don't want to keep paying you just so you can keep changing it.

We need to go back to treating software like it was something you bought in a box off a store shelf and that was it. No recurring payments. No updates. No accounts. No nagging. No telemetry. Just a box of bits that I hand over money for once.

Yes, traditional software also has a cost for maintenance and features but users could opt out to not pay for them.

I personally rather to either pay per usage, pay for the full year if it's valuable enough to purchase a license for desktop / mobile apps if possible.

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…

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.

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.

Personally I work at a FAANG adjacent big tech company and dream of quitting every fucking second

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.

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.

Further the issue of exposing all your sensitive data to the AI provider.

I’m not surprised the business is struggling - the author hasn’t found a deep enough problem to solve.

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.

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…”

Everyone has a job with a remit and responsibilities. The role of business leadership and management is to make sure the people and remits operate harmoniously to efficiently reach a common goal.

If some people doing their job well decimates another persons job in the same company, that is 100% the fault of the people managing the jobs, not the people doing the jobs.

A job is not an object, a human is not a machine.

You can be another Nero. Come, burn everything down, and leave, or you can be a compassionate with people, talk with everyone, win minds and hearts and accomplish the same thing.

I worked with both kinds of people, clashed with the first even if we were trying to accomplish the same thing, and cooperated with the latter, even if what they wanted is complete opposite of my views.

I behaved that way, because the former one wanted to do the job, get the praise, and get me fired in the process, but the latter one showed me an outside perspective I failed to see, so I agreed and did whatever I can to fix what's broken from outside perspective.

If the first person didn't see me as a threat to themselves, I'd have pushed them to the place they wanted to be, because I'm no ladder climber, but they don't work with us anymore.

Of course there will be stubborn people to do whatever they can to protect themselves, their position or whatnot, but they are rare and generally have other motivations, from my experience.

there are so many people just phoning it in, or even worse being (both) socially and technically incompetent.

by what OP said we don't know their usual M.O., and just because they said they can and do cut through bullshit in meetings, does not mean that they are not compassionate with the people who are aligned with the outcome that management wants.

simply because OP does not achieve world peace we shouldn't assume they go around sloshing kerosene everywhere and giving out free hand-held sparklers.

...

and probably the reality is much more mundane and boring. usually technical people have absolutely reasonable concerns , yet most of the time they misjudge the cost-benefit ratios.

I believe that how someone presents themselves tell a lot about them.

Are there "technical people" who will volunteer to refactor an Excel monstersheet instead of suggesting a complete rewrite in whatever framework with an ORM?

Are you perchance paid by number of distinct metaphors used?

>The amount of unqualified ego tripping in tech has become astounding and can serve as a powerful ally for humble wizards.

beautiful, funniest thing i've read all 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"

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.

A lot of the Internet says otherwise — especially these days. Care to share your sauce?

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.

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.

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