Anthony Indraus

4 things that kill prototypes when they hit real users: - Caching layers you didn't plan for - Database queries that worked fine at scale of 10 - Error handling nobody thought about - Monitoring that doesn't exist yet Prototype ≠ production. Most people learn this too late.

You can't polish your way out of being wrong. Some product teams are so focused on response time they've forgotten to ask if the response is actually correct. Speed without accuracy is just expensive noise.

Shipping is the easy part now. Anyone can build. The real skill is getting someone who has no reason to listen to actually pay attention. That's the real product.

I think people assume confidence means you know what you're doing. For me it just means I'm comfortable not knowing and willing to figure it out anyway.

You can tell when someone's building vs performing. The builder spends months on things that look like nothing. The performer optimises for what looks impressive. One compounds, one doesn't.

The best product people I know have this weird habit of just... throwing half-formed thoughts into code and seeing what sticks. Most of it gets deleted. But sometimes you stumble onto something real just by moving fast enough.

@rxhit05 The thing is, I've seen founders build in a vacuum for 18 months thinking distribution would solve itself. By then they're out of runway and too tired to sell.

Most SaaS founders die at $0 MRR. Not because the product was bad. Because nobody knew it existed.

@thdxr the spreadsheet demo is where 90% of them lose me. if your ai can't articulate why it matters before showing me a use case, the tech probably isn't that interesting.

every single ai product: "we've built revolutionary tech that changes your life using artificial intelligence so say you have a spreadsheet...."

@jrfarr "Consistency compounds harder than intensity", that's the one. Most people I know who built something real didn't have a viral moment; they just showed up every week for years.

Distribution is no longer optional. Everyone wants a growth hack or a viral loop, but most of the growth at Lemon Squeezy came from doing a lot of small things for years. Some practical things that worked for us: 1. Shipping constantly We created “Lemon Drops” and every Friday, we shipped something. Sometimes big, sometimes tiny. But every single week we had: >tweets >a blog post >a product release >a changelog update >screenshots/videos >customer conversations This mattered way more than trying to engineer one giant launch every 6 months. 2. Turning product work into distribution Every feature became content. Every integration became content. Every customer problem became content. We stopped thinking: “How do we market this?” And started thinking: “How do we package the work we’re already doing into something discoverable?” 3. Building evergreen content loops We built “Wedges” and gave it away for free (open source). It wasn’t directly monetized, but it gave us something useful to share constantly. And designers and developers loved it. People tweeted it. It ranked on Google, and it introduced people to the brand. Over time, it became a flywheel. A lot of good distribution is just creating assets that keep working long after you publish them. 4. Obsessing over onboarding I think founders massively underestimate this. Reducing friction is distribution. Every extra step in onboarding kills word of mouth. We spent a huge amount of time improving signup flow, activation, dashboards, copywriting, error states, emails, and all the boring stuff. Growth gets easier when people actually make it through the front door. 5. Making docs part of the product Our docs drove an insane amount of traffic. Not because we “did SEO” but because we answered real questions developers were searching for. Most company docs sound vanilla. We tried to make ours actually helpful. Distribution increasingly comes from being useful at scale. 6. Integrations everywhere Every integration unlocked another ecosystem. Another search surface. Another community. Integrations are underrated forms of distribution because they borrow trust from existing platforms. 7. Founder-led content I think I had <1,000 followers when I started Lemon Squeezy, but people trust people more than logos. Especially now. I posted constantly. >lessons >launches >podcasts >screenshots >customer stories >product thoughts Founders underestimate how much simply showing up every day matters. 8. Customer support as marketing Early on, support was one of our biggest growth channels. Answering fast and being human matter. People remember how you make them feel when something breaks. If you follow me, you know I still live by this, and I've carried this mentality into my role(s) at Stripe. Support builds trust faster than ads ever will. 9. Screenshots matter more than people think It sounds silly, but it’s true. Products that look good spread easier. People tweet screenshots, and good design is distribution. 10. Launching over and over again We never really stopped launching. >every feature = launch >every milestone = launch >every integration = launch >every partnership = launch Not in an annoying way. We just consistently stayed in motion. The internet rewards momentum. 11. Building in public before it was cool We talked openly about numbers, growth, problems, product decisions, and lessons learned. Transparency created trust, and that trust created distribution. 12. Creating systems instead of random bursts of marketing This is probably the biggest thing. Most startups market in bursts. They build towards one big launch and then disappear for 3 months. We built systems: >weekly emails >weekly content >weekly launches >weekly improvements >weekly customer conversations Consistency compounds harder than intensity. Product still matters deeply, but a good product alone is rarely enough anymore. Looking back, almost none of this was one giant breakthrough moment. It was thousands of small reps stacked on top of each other for years. That’s what compounds.

I think most side projects die here. They're not products yet, they're demos. 1. Only impressive when you pitch it 2. Breaks on anything unexpected 3. Users can't figure it out alone 4. Nobody comes back after day one

AI didn't lower the barrier to building. It lowered the barrier to starting. Those aren't the same thing. Most people confuse a prototype for a product and a product for a business.

4/ I had to learn that marketing isn't separate from the product. It's part of the product. How people discover it, understand it, choose it, that's the experience.

I built something I was genuinely proud of. Launched it. Then waited for people to care. They didn't.

3/ The uncomfortable part wasn't that it failed. It was realizing nobody owed me attention. Not my audience, not the market, not the algorithm.

2/ I think I expected the product to speak for itself. Like quality was a distribution channel. It's not.

@yashhq_22 the solo founder also has no payroll to meet. which changes what "outlast" means. funded startups die fast, yeah, but they die trying to scale. solo founders die trying to survive. different games.

hot take: the solo founder who posts daily will outlast the funded startup. distribution compounds.

There's this moment right after you ship where you feel like a genius. Then production starts smoking and you remember you're just figuring it out like everyone else.

Every job I thought was a detour ended up being essential. Web design taught me marketing. Marketing taught me product. Product taught me how to think about systems. You can't see it while you're in it.

@jackfriks the gap between "i built something" and "people actually use it" is still the same width it was five years ago. tools just moved the starting line.

its pretty funny that you can build an MVP in a day now without knowing how to code but still the execution of a good idea is not any easier than it was last year getting results still takes something else, but im not sure what it is... maybe its time and raw effort?

@iishaparekh 💯 and it will continue getting harder when the attention landscape is constantly changing

marketing is 100x harder than programming