Built 2 SaaS and got 0 customers. Here's how.

Great breakdown on the quality vs volume debate. I've been doing cold outreach for GoldBean API (our AI API marketplace) and personally found that mentioning a specific pain point in the first sentence gets 3x more replies than any template. The hardest part isn't sending messages — it's finding the right trigger. For developer tools, 'API pricing transparency' has been my best opener because every dev has been burned by hidden fees.

Hey John, noticed you’re building AI tools.
I analyzed 500+ AI signals from GitHub, Product Hunt and Reddit this week.
One interesting pattern: AI agents for customer support are showing repeated demand signals.
Thought this might be relevant to your work.

Im in the same boat

Brutal honesty. Most builders focus too much on the product software rather than the raw market psychology. Keep executing. 👑🏁

What makes this post valuable is that the real failure wasn’t building the wrong product—it was confusing product progress with business progress.

Steps 5, 6, and 9 are especially important. You can spend months improving UX, adding features, and polishing the product while avoiding the one activity that gives you the clearest signal: getting it in front of the right people and having uncomfortable conversations with potential customers.

I also like the final point about blaming the idea. “Nobody wants this” and “nobody knows this exists” are two completely different problems, but they can look identical when you're staring at an empty analytics dashboard.

The biggest takeaway for me: validation, distribution, and product development shouldn't happen in separate phases. They need to happen together from day one. Great post—and congrats on turning those painful lessons into a successful second act.

Step 4 and Step 9 are uncomfortably close to home right now. Spent weeks getting the positioning, the SEO, the affiliate program all lined up before publishing a single piece of content anywhere the people who might actually buy this stuff hang out. Zero sales so far — which stings less as a verdict on the idea and more as confirmation that "built it properly" and "anyone knows it exists" are two completely different projects.

The thing I keep circling back to from these comments: the "did anyone complain about this before you built it" test. In my case, yes — the failure modes I built around are ones I'd watched play out repeatedly in enterprise programmes. So I'm fairly confident it's a distribution problem, not a nonexistent-problem problem. Doesn't make distribution any less hard to actually solve, though.

Step 10 jumps out at me.

It's easy to say "the idea didn't work" when it's really too few of the right people ever saw it. But the opposite can happen too. You can keep telling yourself it's a distribution problem long after the market gives a weak signal.

I've run into this with products I've built. The difficult part wasn't just doing more marketing. It was deciding what evidence would actually convince me to keep going, change the approach, or stop.

Looking back at the two SaaS attempts, was there a point where you could have distinguished "bad idea" from "not enough distribution" earlier? What signal would you pay attention to now?

The “marketing starts after launch” point is probably the biggest one here. I’ve seen this happen a lot.

Getting the product in front of the right people early can save a lot of time. You learn what people actually care about, what makes them convert, and what needs fixing before spending too much time building.

Otherwise, you can end up with a polished product that nobody knows about or needs.

Step 6 hit hardest, mistaking building for progress is basically the whole indie hacker trap. I did the exact same thing with my first app, ignored every signal for months. Weirdly comforting seeing it laid out this bluntly, most of us dont admit it this clearly.

Sharing these post-mortems is super valuable! Are you planning to completely abandon these projects, or pass them on to someone with a marketing background? I built Digimarket as a 0% commission marketplace with an AI valuation tool specifically for indie hackers wanting to offload side projects.

This hit hard.

I’m a factory worker in Italy and I launched my first Android app a month ago, built in my spare time. It solves a problem I had personally: knowing how much I can safely spend before the next paycheck, without having to calculate everything manually.

The app works for me, but getting it in front of people who don’t already know me has been much harder than building it. I tried a small launch, X, Hacker News, local press outreach and communities; so far, almost all installs are from people I know.

I’m trying to learn the difference between “the product isn’t useful” and “not enough of the right people have seen it yet.”

Thanks for sharing the honest version of the journey.

It's all sarcasm on the surface, but there's a pattern there: most of these errors are the result of optimizing for building and not learning. Probably the largest point is the “build the real version first”.

The more advanced the teams, permissions, integrations, analytics, and polish, the more it might seem like progress and the more it might make one wonder, “Does anyone really want this?” It's helpful to differentiate between product risk and implementation risk. If no one would like the product, then quicker solution to the technical problems doesn't help much.

That's also noteworthy. If a thing fails to gain any momentum, “bad idea” might be one reason. Sometimes the product just failed to reach a large enough number of the appropriate target audience to make a significant splash.

As an angel I can usually tell within five minutes which founders built in silence: they ask how to find customers, while the ones who sold first ask how to build faster. Henson Group only worked because clients hired me before I had anything resembling a product. Selling first is the cheapest validation there is.

Haha, "Step 2: Build the 'real' version first" hit way too close to home.

It’s so easy to fall into the trap of thinking that more features = more value, while ignoring the fact that nobody actually needs the product yet. I especially love the point about "Mistaking building for progress." It feels productive, but it’s often just procrastination in disguise.

Thanks for the reality check (and the laugh). The "PS" gives me hope that we can all learn from these painful mistakes!

The intent-based targeting point is really valuable. Reaching out when someone has already shown interest in the problem feels much more sustainable than simply increasing connection volume. Curious how you track whether that intent signal actually translates into better conversations over time?

Heheh thats really true, i agree with your points. Also when someone gives negative feedback about your product it feels demotivate ,self doubt too but we need to move on, one single feedback doesnt decide your product i believe.

haha that's really funny and honest. We have all been in this loop recently. But let's be honest, even before starting to build, doing a small market search is not as pretty straight forward in my opinion. We all have social media accounts, but building that engaged audience is not as simple as it seems, at least from my experience.

I treaded my two Saas as; first tools that solves my own problems, trying to make my life easier in the future. Also, they are something valuable for my portfolio, something I'm proud of. Of course, getting traction and having high conversion rates is the real deal when talking about Saas, but there is still that engineer inside me who wants to play around and release stuff and try something new, rather than just consuming.

Have been stuck in the loop of step 1 and step 6. Is it confidence lacking to go out and tell the people that i have built something cool while being uncool?

This is genuinely funny and a good mirror. Going through it point by point, I think I've actually avoided most of these, 113 days of daily posting (opposite of building in silence), actively working on not compulsively refreshing analytics, doing outreach and SEO from early on rather than waiting for some perfect launch moment.

But a couple hit close to home. I did launch with a genuinely full feature set instead of the leanest possible MVP,not sure if that was wrong for what I'm building, but it's a real match to step 2. And the closing line, blaming the idea instead of the fact that nobody knew the product existed, that's basically been my entire visibility problem for the last few months, not a validation problem, a "how does anyone find this" problem.

Congrats on the sold company and the $3K MRR on the current one, would genuinely be curious what changed between the failed attempts and this one, beyond just the mistakes listed here.

Step 2 and Step 9 both hit close to home. I built a landlord compliance tool (CertMinder) after getting fed up tracking gas safety/EICR/EPC dates across spreadsheets, and the pull to keep adding "real version" features before showing it to a single other landlord was strong.

What actually shifted things for me was treating testing as the finish line, not the launch: multiple test accounts, deliberately breaking it with messy real-world data (expired-yesterday certs, HMOs needing five different licence types) before I'd trust it enough to tell another landlord to rely on it. Cheap insurance against exactly the "built it in silence, nobody came" outcome you're describing.

Though I'll be honest, I probably leaned on thorough testing partly because it felt safer than doing Step 9 properly (i.e. actually talking to people before I was "ready"). Two different disciplines, and I only had one of them down early on.

Step 6 hits hard. Building feels productive, talking to customers is scary 😂

good going

Step 9 is probably the most dangerous one. It’s easy to convince yourself that polishing the product is productive,when talking to potential customers would teach you more in a day than another week of building.Learned that lesson the hard way 😅

Step 5 and 9 hit uncomfortably close. Built CheCeno in near silence, told myself I’d “figure out marketing after launch.” After was a lot further away than I expected.

The part I didn’t expect: even my AI coding assistant never flagged it. Answered every technical question well, never once asked why I was staying quiet about positioning. Turns out that’s on me to ask, not on the tool to notice.

Congrats on the exit + the current one — genuinely curious what changed your relationship with Step 9 specifically. Did marketing start earlier this time, or did something else shift?

Brutal honesty. Most builders focus too much on the product software rather than the raw market psychology. Keep executing. 👑🏁

For step 7, apart from the vision itself, what other signals helped you trust that it was the right direction?

If not completely, sums up quite a bit on my journey in building FileLever :D

Two rounds of “built it, nobody came” is a rough but common rite of passage. What’s different about how you’re approaching validation this time versus the first two?

Step 6 is the one that got me hardest. "Every new feature feels productive. Every conversation with a customer feels uncomfortable." That's not a bias you can decide to fix. It's structural. Building gives you feedback every 20 minutes. Talking to customers gives you feedback you can't always act on immediately. So your brain correctly concludes that building is higher-reward in the short term.

The only thing that broke the loop for me was a rule I had to make external: no new features until I'd spoken to 3 users in the last 7 days. Not "talked to users recently." Exactly 3, exactly 7 days. The specificity was the point. It's too easy to fudge a vague commitment.

The bit about selling a company is also worth noting. Most people don't write about the failures that preceded the exit. This is a more honest version of that story than the case study format usually allows.

We have all been here and keep doing it over and over again. Congrats on the successes in there as well.

I feel personally attacked by step 5 and step 9.

I spent three months building a Windows security tool. Ransomware protection, 32 detection rules, 32 hardening rules. Felt productive the whole time — shipping features, fixing bugs, making it solid. Then I launched it into complete silence. Zero users. Zero traffic.

Took me a while to realize that building the product was the easy part. Getting anyone to actually use it? That's the real challenge. Still figuring that part out.

Thanks for writing this. Made me feel a little less alone in the "built something and nobody came" club.

This really hits home. I’ve spent a lot of time building my AI tools platform, and now I’m realizing that building the product was only half the job. Getting real users and feedback is the harder part. I’m starting to focus more on talking to users instead of just adding features. Thanks for sharing this!

Nice

I hear you loud and clear! I was taught that same lesson today. Six months - three platforms - $0. Why? Because I built first, then offered a new tool to customers that hit them out of the blue. Some were shocked, some didn't even care, others copied and built similar on their own. LessonS learned.

are u talking about me

This hits home 😄. I'm currently building my own SaaS, and one of the biggest things I'm learning is that building the product is actually the easier part. Getting people to discover it, talk about it, and eventually become customers is a completely different challenge.

I'm building Postlefy, an AI-powered social media scheduling and automation tool, and I'm realizing that marketing and talking to potential users need to happen alongside development — not after everything feels "perfect."

Definitely learning this the hard way too.

That is nice

The part I’d add to this post is that “build vs. market” can still hide a conversion problem. You can talk to users, validate the problem, and still end up with a SaaS that people understand but don’t feel compelled to buy. The missing layer is often the translation from “this problem exists” to “this product is the obvious solution for me.” I’d be curious, looking back at those two SaaS attempts, whether the people you spoke to could clearly articulate why your product was different from their current workaround—or whether the validation stopped at confirming the pain.

This reads like a joke until you score yourself against it. I got to zero customers with a fully produced course platform, videos and all, so number 5 and number 9 hit hardest. Building genuinely feels like progress while you're doing it, that's the trap. The fix I've settled on is embarrassingly small: a daily 30 minute block where I'm only allowed to talk to people, not touch the product.

This hit pretty close to home. I’m currently building my own SaaS ecosystem, and one of the biggest lessons I’m learning is that building more doesn’t necessarily mean moving forward. It’s surprisingly easy to spend weeks improving the product because you know what you want it to become, while avoiding the uncomfortable part: putting it in front of people, listening to them, and finding out whether they actually care. I’ve definitely caught myself doing some of these things already, especially treating features and polishing as progress. Now I’m trying to reverse that mindset: ship earlier, talk to users earlier, and let real feedback influence what I build next. The “build in complete silence” point especially hit me. Thanks for sharing the lessons from the failures rather than just the success story. Those are usually much more useful.

The uncomfortable question that helped me most in this spot: did anyone complain about this problem before you built the tool? Not "would you use this" — people say yes to be polite — but spontaneous, public complaining: forum threads, app reviews, support rants about the incumbent. If that trail exists, it's a distribution problem and it's fixable. If it doesn't, no marketing fixes it. Worth doing that autopsy honestly before spending another month on growth tactics.

nice

nice

Thanks for sharing openly — traction is the hardest part. I’m building Finsight AI, and I’ve found customer discovery is more important than features early on. Curious, what would you do differently now to validate demand faster?

This hit home. I’m building my first SaaS right now, and one of the biggest things I’m trying to avoid is hiding behind “building.”

My instinct is still to improve the product, polish the brand, add features, and tell myself I’ll focus on distribution once it’s ready.

But I’ve realized that’s exactly how you can spend months building something nobody asked for.

With FollowReach, I’m trying to take the opposite approach: talk to people, share what I’m learning, publish the problems I’m seeing, and use real conversations to shape the product.

Still early, still learning, and definitely easier said than done but I think getting comfortable with customer conversations is probably more valuable than another week of polishing the product.

The “build in complete silence” and “mistake building for progress” points especially hit home. 😅