I don't doubt that you're absolutely right, but this bit strikes me as hard. I've struggled to find one business owner with a problem they perceive, let alone 10 with commitments to buying a solution for the same problem.

Ironically, it strikes me that there's a gap here for a business to solve; connecting entrepreneurial devs with business owners aware of problems.

Draw the funnel diagram, with numbers if necessary. Having talked to many devs who believe similar, I think the most actionable advice is likely “Organize your next N weeks to talk to many, many more business owners than your last N weeks.”

To give a counter-example, I did consulting for a decade before going off to do my thing, so I have a lot of personal contacts and I'm familiar with a lot of problems, both general and specific, that businesses struggle with, and turning those into products has been relatively straight-forward. But most solo devs don't have this background, so they need more specific and actionable advice to build their funnel.

There are like eight zillion guides to doing this online, and unless you're willing to spam people, it always boils down to: Grind it out. Start making cold pitches, and hustle for as many warm intros as possible.

Learn sales y'all.

They want “interesting problems” instead of just trying to create products that someone is willing to pay for.

After over two decades, despite all of its attempts, Google has yet to do anything successful that wasn’t advertising based.

No Android doesn’t count. It came out in the Oracle trial that Google had only made $24 billion in profit on Android from its inception until the beginning of the trial. I’m the meantime they pay Apple $8 billion a year to be the default search engine on iOS devices. Apple has madd more in mobile from Google than Google has made from Android.

What I called technically interesting is "working on a large scale product", which for you seems to have implied "work on a new problem", but for me it's exactly the opposite. The kinds of large scale projects I find the most technically interesting are long into their growth curve.

Personally, I find boring to be quite exciting - but perhaps not if the only thing you're looking at is the technical solution. Building a small business involves figuring out how to solve the mundane but important thing, but then you get to figure out how to sell it, how much to charge, how to do customer outreach, do support, turn the business into a repeatable process, etc. Building a business out of it is quite challenging and fun and the technical solution is frankly a small part of it.

Many solo-preneur types with technical backgrounds are way over-indexed on the technical aspect and want to treat it like their last technical job, but without a boss, which leads to a lot of disappointment and frustration.

i'm kinda in that situation now, and got to play all the roles, like how we increased our conversation ratio by 50% with me thinking like a sales person (hint, don't over-complicate your pricing page), and those things can get you exited for some time and the money also, i just don's see it long term. i would rather playing with something that i love to do.

I am lacking the domain experience to even have the first conversation, and it feels like paying people to ask "so tell me everything that is wrong with your business" is the wrong way to do it.

I know PG mentions "solve problems you have yourself", but I am not a business owner. I'm a software engineer.

That video you linked earlier - Designing the Ideal Bootstrapped Business - was incredible and feels like the right move once you have that idea. But what about finding that idea?

Most business owners tend to be experts in their specific domain, where they have already used their knowledge and expertise to great advantage. But running a business entails more than just handling the problem domain itself. There's an entire category of tasks and responsibilities under the "business administration" umbrella that the business owner will not be an expert in, and will probably despise dealing with because it takes them away from the fun stuff that is their expertise. That's a good area to focus your gaze on.

For instance, the accounting clerk may be spending 4+ hours every day manually typing invoices into two different systems, but to the business owner this is probably perceived as a normal and expected course of business. The accounting clerk is unlikely to complain about it themselves either, since they are getting paid to do it (remember that Sinclair quote). But to you as a software engineer, double data entry is very obviously a red flag, and if you care enough about it you can do a deep dive and see if it can be eliminated using automation, and present it as a cost-saving solution.

Generally speaking, nobody is going to hand you a written list of their problems on a silver platter, and even if they do, the problems they have identified will be so general and vague (e.g. "we have a lot of inefficiencies in our accounting department") that you won't be able to simply go home and start hacking away at them. So you need to use a methodology to start peeling off the layers of the onion, so to speak. And that always entails follow-up meetings and learning other software systems and familiarizing yourself with various business practices.

At some point, you will come to the realization that you no longer view yourself as a "software engineer". Rather, you are a problem solver, and writing code is simply one of the many skills you possess. That's when you'll know you're on the right track.

I was in Software for a couple of years and I only saw software/IT problems. But these problems usually had a variety of solutions and only some glue was required.

Now I've been in the natural stone industry for 4 years and I can't count the number of problems to which the solution is "add more people". So much data entry and data extraction from PDFs <-> ERP/CRM/Other software. 100s of man-hours spent on something that could be done with proper data formats and simple automation.

I personally believe that software engineers are SORELY lacking on all other industries besides software. We need more software engineers venturing out into other industries and identifying and solving problems.

this seems to line up with my experience well. But I don't have a good line of sight for me to actually experience those businesses outside of taking a non-SWE job and going from a well-compensated expert to barely-paid newb. Spending years learning a specific industry in the hopes that I can turn my former skills into a viable business seems like the wrong approach in several axes. I know VC groups (sometimes) have entire departments whose purpose is to understand other industries... it seems counterproductive for individuals to try to and go this route. Thoughts?

Some potential paths can be:

1. SWE @ SW company -> "BA" @ non SW company or whatever term is used for generic-problem-solver in that industry.

2. Part-time hourly work in another industry. This is fairly easy through temp agencies.

3. Apprenticeship in another industry (perhaps in more hands-on industries).

Some are hard to get, others don't pay as well. But to find gold, I'd say some level of hardship is required, and this, in my view, is a bulletproof way of finding that gold.

Whether you're able to dig it out and then able to sell it at a profit, is totally another question.

This sounds like a more brilliant idea than #2 in a list in a long thread.

Temp agencies are where businesses turn to when they have a reasonably simple problem and they are willing to pay to solve it.

Out of curiosity, what's your role in the space you're in now? Are you solving these data extraction problems?

I don't have a fully working solution, just started testing with specific documents (steamship lines notifications) to see if it's a fit.

I think it'll be slow road for each department I want to tackle - logistics, accounting, purchasing, etc. Every department has this issue, and I'm sure products could be built to serve those niches.

Why don't you start from your own problems? I bet you have some of them and you are not alone definitely.

In my experience business owners tend to talk about problems that are impossible to solve, or problems that are trivial.

Typically, when you actually find problem worth solving, it is not merely a software. You are looking at whole processes that need to be changed. You will change how people work. And when you actually try to present the new solution to the organisation, you will face resistance. People don't like change.

Some could say that it is the business owners responsibility to push the change. But when they give up, they stop paying you.

Anyways, the advice "just find a problem" feels to come from people who either read too many books and never tried, or from people who got lucky the first try.

This struggle is real. People also don't understand that software development is a process of continuous optimisation, learning and improvement. They won't accept a good solution only a perfect one.

No one really wants to deal with a saas when a one time purchase is available. Things like free tiers, lockins prevent us from moving until something happens (something always happens from big price increases to service changes or shutdowns changes to terms, etc).

Biggest problems are cost and uncertainity and losing a sense of control.

A saas to a customer is run like a fly by store selling stolen goods out of the back of a truck. Even though this person has been selling at this same spot every friday you know this can't go on forever so when he doesn't show up you are not surprised. When a saas closes down / changes terms / increases prices suddenly you are not surprised.

It works with some industries that have a high rate of change but if you plan on having a stable business you want to avoid saases

Every "tech" company is not actually selling tech, but rather is selling something tangible, a product of sorts. Software developers are selling their labor.

If you want to find out the scalable product or solution in a particular business or domain, I think you really need to be a part of that industry or have someone who is.

Otherwise, even if you end up developing a SaaS that satisfies the needs of 100 customers, you won't really ever truly be able to scale. So my suggestion is start talking to friends and family in other industries.

Beyond that, every business owner kvetching on Twitter. Problems are everywhere!

Certainly can be very hard to get 10 business owners with the same problem to all line up and pay thousands for your software, but problems are easy to come by.

Edit: filtering for problems solvable by any particular solo developer's skillset in less than 2 years is another challenge.

I have a backlog of like 200+ viable new product / feature ideas that help solve some business problem or improve the efficiency of activities people do with financial transactions involved. Many are too small to be stand alone for sure, but I do not believe million dollar ideas are grown in a silo

But as an example I also get a lot of fluff ideas for video game mechanics. For multiplayer games, the ability to tag another player for tracking. So you can pull up a menu and see players you've tagged and their recent performance and such or create your own 'achievements' and see what players have accomplished them (with necessary privacy options included of course). The primary use-cases being clan recruitment and pro-player analysis.

Random product idea example: A renting service for size-adjustable tables. So you can test out what table size fits best in your space. counter-height vs standard, square vs skinny, size A in room 1 or size B in room 2, etc. probably too small a market on this idea though since itd only be worth it before expensive table purchases

https://www.indiehackers.com/podcast/136-dave-sims-of-floify

This. Why hasn't this gap been filled yet?

>This. Why hasn't this gap been filled yet?

It's called Upwork.

No seriously, hear me out. When a business owner has a problem they perceive could be solved with technology, they create a job or they go on job posting sites seeking either an employee or a contractor to try and solve their problem.

Where they get it wrong is that they frequently have unrealistic expectations about what it will cost and how long it will take. They have no idea what skills they should look for and who to trust.

Is there a market mismatch here? Absolutely, but you'll have to find a way above the fray of recruiting sites as that's really the state of the market.

The freelancer sites are just a starting point, they answer the question of, "what are some things business owners want".

Therein lies the problem: They know lawyers, doctors, accountants, heck even plumbers are going to be expensive because they are educated / trained and/or they get you out of an identifiable tight spot (the IRS isn't happy, water is coming into the house, I'm having a heart attack, I'm getting sued, etc).

Software developers? Outside the FAANGs technical people are viewed as fungible. Many, many small businesses are grossly undercapitalized such that a business owner might very well be paying that technical person a lot more than they pay themselves. Likewise, technical roles are typically compensated well above many other clerical / field / service / business type roles such that a business owner might not have ever paid so much to any other single employee.

Fungible does not mean low-paying. Plumbers are mostly viewed as fungible, however expensive they may be.

Or something like that!

I used to have an Upwork account, to try and swing some cash on the side. Most of those jobs are paid far out of proportion to what they are asking people to do. "Facebook but for XYZ", is something I'm pretty sure I've seen there, several times.

I suspect that you can get a lot more leverage visiting your local business organization watering holes...

That might be true, but it doesn't negate that there are a lot of business owners there doing exactly what I stated.

Someone who solves or even slightly mitigates either of these issues will have a viable business. I know that for the "ensure the dev is competent" side of things there are many consulting companies that live and die by their reputations, but generally these don't usually scale up beyond a certain size, and when they do their reputation deteriorates. I am not aware of any businesses that attempt to solve from the flip side and weed out problem clients (if anyone is aware of such a service, I would be interested).

As if that wasn't enough of a challenge, if you do manage to connect quality devs to quality clients, there's the issue of sufficiently monetizing that relationship: you can charge a finder's fee or similar when you first connect them, but once a client finds a dev they like they will probably stick with that dev. So in that business you can lose business because you match people too poorly (and they leave because you can't help them) and you can lose business because you match people too well (and they leave because they've now found somebody they trust so they don't need your services anymore).

[1] https://en.m.wikipedia.org/wiki/The_Market_for_Lemons

Then on the other side of the scenario, the dev has to be logistically available to working on the project, be capable in both the tech and soft skills needed, care about it, have the proper motivation (financial and mental), and network.

The hypothetical business would solve the networking issue, but not the rest. But I think these relationships are built in a more decentralized way, though chance encounters, mutual friends, and cold connections. Sort of like dating. (Kind of is because this relationship seems like one of cofounders.)

Or I could be completely wrong and there is a stealth startup out there ready to shake things up.

1. Know someone at the company or a key advisor 2. Spend time together 3. Hear "we think we need help but we're not sure where" 4. Respond relatively insightful with 2 or 3 things they could do that are valuable and force rank them 5. Get feedback 6. Price it. Explain value. 7. Deliver

In my eyes if you're helping smaller companies sub say 1000 employees the value you bring is in knowing what to do. And telling them. They need you to understand the universe of options, what you've seen at other places (i.e. experience) and to weigh in on what works. They largely have no idea how to move the boulder otherwise you would not be there.

This is why I faced finding solve-able problems much easier as a consultant because once you're embedded in an organization you see their problems, and perhaps you can ask the right questions and tease out a possible sale-able solve-able solution.

"... I outsourced much of the writing. That cost me more than it should have because I knew nothing about hiring and managing writers, but the experience taught me a lot ..."

"... no love for Xero ..."

As an entrepreneurial dev who is also a business owner, it sounds like he may have some problems there to solve!

Not sure if those are viable, just pointing it out. The easiest problems to solve are your own or at least ones you encounter yourself. Doing some consulting may help to see problems in business or industries that are not your own.

To be quite honest: He's spending all day playing.

That's a position I envy quite a bit.

You can easily amass 500k-1M and just coast on 40k/year in expenses for a long time. If it's something you want... a lot of people want kids, or to live in high cost areas, or to accumulate wealth to pass on, or have dependents such as elderly parents, or enjoy money a lot.

With the usual caveat, "if you are working for a big tech company in the bay area".

In London, starting on 30k and rising to 80k after ten years, not so easily.

I’m fortunate enough to be in this situation, where if I go somewhere cheap and keep my expenses down I can sustain myself financially for a long time, maybe indefinitely. And I’ve though about leaving my salaried job to do my own thing many times. I do worry though that the lack of financial pressure will just leave me lazy and “playing around” instead of focusing on building a financially viable business.

Anyone else have experience with this?

The one thing I'd say, is that it's critical to know whether you have a strong internal drive / self-motivation. If you don't, it can be a dangerous context to put yourself into (dangerous in the financial or time waste sense). If you don't have that internal always-on motivation, it's super important to have a firm plan and object/thing/business idea that you're going to be pursuing before you quit. Start at least drawing up some sort of minimum framework before you quit, laying out what you're going to do, how you'll spend your time, what type of lifestyle businesses you're interested in, plot everything out financially as much as possible and so on. Good advanced planning will save the day and help prevent catastrophy as the road gets challenging. It's a new type of 9 to 5 job, although it's more likely to be a 50-80 hour job getting started. Without financial pressures, you have to either naturally self-motivate strongly so you don't constantly veer off (get lazy, get distracted, get bored, and such), or follow a fairly strict plan that you've laid out for yourself ahead of time to keep yourself sane, productive and making progress.

Harder if you're in a higher cost location but you can definitely live off $60K even in SF if you're single.

In a sense, it's cool, cuz like, you're helping people in the world out for less than the cost to provide that service. But in terms of recommended reading material for people hoping to replicate the situation without an infinite runway, selection bias is a definite concern.

The argument would probably be that he's lucky to be in the position he's in. But presumably he earned the money he has from working at Google. So either dismiss his entire situation by saying that anything he tries to achieve from now on is just "playing around" because he already has enough money, or respect him for trying to continue to improve his life (even if it's from a good place to start).

I don't think it does, but looked at purely through the lense of 'how to be a solo entrepreneur' maybe discount advise from folks who can afford to be wrong for years.

I want to distance myself from the 'you should just work for google' folks. I totally respect the idea that using your savings to build the services you want from society. My position is that by default, postings on HN are viewed as recommendations to others. And yes there are other reasons to write, even if nobody else is reading it. But with zero qualifiers (Show HN, Ask HN, etc), my expectation from HN front page material is that people should read and learn from the material.

Anyone hoping to learn from the post for their own ambitions needs to be able to separate the typical from atypical in the author's post. The author is forthright that he's funding this via savings and I think it's reasonable to think he can afford to do so in perpetuity -- again, great for him; I hope to be in a similar boat some day. But like, last year's post describes a cost cutting strategy of buying a house, and most self employed persons find it challenging to get a mortgage without multiple years proof of income.

You make it sound easy! How do I do that outside of not just SF, but the US?

Work for one of the big tech cos outside of US is probably your best bet.

Bag-holding 101, folks.

It looks to me like some people have a well-developed intuition to distinguish between products in "seems like it would be useful to me" category vs. products that have a good chance of actually being commercialized.

I certainly don't. When reading various indie hacker stories I'll be damned if I can predict with reasonable success which ideas have a good chance of taking off and which don't.

I strongly disagree with the perspective here. Cost control isn't the core problem for the author, but every dollar matters at this stage. The difference between "we're losing money every month" and "we're break even" and "we're slightly profitable" is huge when talking to parents and other people in the social circle.

> Your minimum commitment should be north of $1,000 per month; practically nobody can integrate with your API for cheaper than a thousand dollars of engineering time, right.

That's my thinking as well, but haven't had much luck getting companies to accept this kind of pricing. Even though it's all inbound interest (no advertising or outreach) and they're sometimes large companies.

Here are the problems I'm aware of:

1. People generally don't think about the cost of standups or meetings or per hour employee cost. That's just not something typically being considered.

2. They've compared with other APIs or use other APIs/SaaS and are used to <$100/mo. So $1k/mo or even $500/mo is a shock and they decline or disappear. Apparently deciding they don't actually need it that bad or going with their 2nd choice.

How do you overcome these? Do you actually point out that the API is cheap compared to employee cost? Help them justify it to themselves and/or their manager?

There are several paid SaaS apps and / or Slack bots that do this. It’s a viable niche, but has gotten somewhat crowded in the past few years. Earlyish example: https://iDoneThis.com (worked there 2014/2015)

>You could probably talk to business owners with problems, launch an (appropriately priced; hundreds to thousands of dollars per year) product against those problems, build contingent on getting 10 commits to buy, and be at +/- $100k in 12 to 18 months. Many people with less technical and writing ability have done this in e.g. the MicroConf community. If you want the best paint-by-numbers approach to it I've seen, c.f. https://www.youtube.com/watch?v=otbnC2zE2rw&t=2s

Thanks, I'll check out that video!

I hear that advice a lot from people I respect, and I put a lot of effort into that over the past year, but I was never successful.

The biggest problem I ran into was that there's this "disconnect problem." I know I'm a competent developer and can build niche solutions to $10k problems, but I don't know which companies have $10k problems. I can guess at it, but they sometimes don't even realize their $10k problem has a software solution. The companies also don't have incentive to talk to me to explore the possibility if I'm just a developer off the street.

In 2019, I tried to do this with stone quarries[1], sheet metal shops[2], and email copywriters[3], but the first two mostly wouldn't talk to me and the latter didn't seem to have enough opportunity for a niche business.

>I would encourage you, to the maximum extent compatible with your sanity, to prioritize "Will this get me more customers?" over behind-the-scenes investments like CI/CD which are very appropriate to Google but will under no circumstance show up in next year's report as One Of The Most Important Things I Did This Year.

I get a lot of pushback about my love for CI/CD, and it always puzzles me. Is CI/CD seen as difficult or time-consuming? For me, it's such a net positive on my time and mental energy to know that basic functionality works before I push to prod. There have been many times where CI has caught breaking changes that I'd otherwise have to catch by waiting for customers to complain or manually smoke testing my product after every push. And I don't have to do much to set it up - just slap in a Circle CI config.yml and flip a button.

>I would suggest devoting approximately zero cycles to cost control. You don't have a cost problem and no amount of cost control will bend the curve of your current businesses to sustainability. You have a revenue problem. Your desired state in the medium term will make it economically irrational for you to think for more than a minute about a $50 a month SaaS expense; marketing and sales gets you to that desired state, not cost control.

Thanks, I agree completely and you succinctly articulated a feeling I've long struggled to put into words.

I often get pushback about spending O(hundreds) of dollars on non-essential expenses, and it always feels like it's missing the forest for the trees.

[1] https://mtlynch.io/retrospectives/2019/05/#an-app-for-rocks

[2] https://mtlynch.io/retrospectives/2020/01/#sheet-metal-resea...

[3] https://mtlynch.io/retrospectives/2019/06/#taking-on-google-...

For example, I'm currently evaluating warehouse management systems. It appears that everyone in this space runs a SaaS; with ludicrous prices where they want to bill per shipper/ac count. We still plan on being in business in 3 years; and after 3 years we will have paid more than what it would cost to develop this system ourselves. If you are mostly focused on tech it's easy to think that everyone likes the idea of SaaS; but realistically the only people who get excited about that idea are the people cashing the checks every month. I have heard much lamentation and whining about having to 'subscribe' to software instead of just buying it.

I don't think any of these businesses would pay for What Got Done, but I think there are a zillion businesses very geographically close to you that would.

I note in several of your posts that you live in Western MA. You're two hours by car (assuming you miss traffic!), or three hours by train, from Boston. I can think of ten companies off the top of my head that fit the mold Patrick mentions as potential customers for What Got Done, that would not bat an eye at paying the upper limit of prices mentioned. (My current employer being one of them!)

Happy to help you brainstorm potential customers via email - I'm at (my_hn_username)@gmail dot com.

Oh, to clarify, I wasn't trying to sell What Got Done to these industries. I was trying to understand their businesses to see what new thing I could create for them.

That's legit.

I guess the core of my original point was that there's a big regional hub of businesses whose operations you do understand (i.e. SW businesses) that's right at your doorstep. (I'm assuming from your other blog posts you don't know a ton about sheet metal bending, whereas you do know enough to be dangerous with software, but I've been wrong before!)

With the quarry example, you couldn't get them on the phone. How about conferences, meetups, trade shows etc? Furthermore you probably need to search for companies that might _want_ to speak to you: younger owners, companies with less to lose etc.

In the sheet metal example, it sounds like you had already decided what you wanted to build for them before even meeting them (an improved software solution for their shop management apps). Maybe trying to just speak more generally about their processes and difficulties would tease out problems you might not have considered.

I'd return to this approach/phase and give it another go to find a market fit for the ideas before you start doing a line of code.

I think that's sound advice. The strategy I had in mind was courting industries that were physically close to me. I wasn't just calling the businesses, I was showing up at their offices to ask for meetings. My hypothesis was that there are probably plenty of profitable businesses nearby that are disconnected from the tech world, so other software companies aren't approaching them.

To clarify with the sheet metal companies, I didn't go in pitching a replacement to their existing tools. I always opened the conversation by asking them if they had processes that frustrated them, cost them money, or if their existing software failed to meet their needs. But I definitely found it hard sometimes when they asked me what I wanted to build for them and my answer was basically, "I don't know..."

Maybe it's just a matter of picking the right spot on the spectrum. It seems like you want businesses that are niche and disconnected enough from mainstream tech that big players aren't courting them as well. But they can't be so disconnected that they won't talk to you, either.

I think [0] that a better approach, one that’s more likely to lead to a profitable business, is to flip the strategy around. Instead of “Do customer interviews, sniff out a problem, and build a solution”, it can be more effective to find the customers where they hang out (online is ideal), observe the problems they complain about, and crucially observe what they already pay for.

For one, it avoids the natural bias they’ll have while answering questions (they’ll very likely tell you what you want to hear to try to be “helpful”).

For another, it helps you create something they’ll actually want, and pay for, by fitting in with their existing habits and worldviews.

Going into a conversion with the idea of “I need a problem to build a SaaS for” is only one level better than “I need to validate that they want my idea”. The format of the solution is already pre-decided — which is no good if they don’t already pay for SaaS. That’s an uphill battle you definitely don’t want as a solo founder! “If I could just make them see how great this is...” is a rocky start for a business plan.

Choosing an audience you know and already understand is a big advantage. Picking one you aren’t part of is harder.

You don’t need to be new and unique to be profitable. You don’t need to invent a new kind of product or solve a problem that nobody else has solved. All of that just makes it harder :)

[0] I think this because of Amy Hoy and Alex Hillman and their 30x500 class that teaches this approach. Not my own invention :) I’ve had success with it though, and other have too, so I want to give credit where it’s due. They have a lot of great free material at https://stackingthebricks.com

That's tough stuff. I've done the same, and while it feels good when you're able to talk in detail with a business owner, I didn't come away with actionable things usually. There just isn't a good substitute to being embedded in an industry.

Michael, work part-time in a warehouse (as manual labor) handling stone for 3-6 months. Guaranteed you'll come away with more problems than you can solve.

measure how often you're talking vs. they are if the split isn't 80% in their favor you're doing it wrong.

They sometimes do realise their $10k problem has a software solution though. You need to be in the pathway for when they go looking to solve it, not go to them.

Advertise in the white pages as doing custom software development. You'll get plenty of businesses contacting you ready to pay money to solve all sorts of problems not currently addressed by off the shelf solutions. Very rarely are they actually willing to pay anywhere near the total cost to build it, but enough to make building it worth it to you if you can get a couple other customers on board.

I deal with stone factories/quarries everyday (but mostly in Brazil and India)!

Let me know if you're still interested in this space.

But my impression of the author was that they are well off already and is just trying to have fun while also possibly earning money down the road. In other words, they are trying to follow their passion.

Even if he's not explicit about it, every lifestyle decision I'm getting from these blog posts is that OP has effectively F.I.R.E'd and these projects are mostly for fun.

I'm curious why you use this wording. It comes of as slightly brash, and to many, suggests that you shouldn't worry about costs. In my experience have to be "penny wise, pound foolish" making reasonable efforts to manage cost.

You can pile on loads of tech debt which matters little at this scale (e.g. your CI/CD example), but when you are small the costs can bite hard. You quickly limit your choices and end up spending time looking for funding, if you don't have some sort of reasonable constraint.

Can you elaborate on this? Is it just because of the inconsistency in revenue or is there some technical challenge I haven't considered?

I'm currently developing a usage-based API SaaS as a solo developer so this in particular really piqued my curiosity.