Get live statistics and analysis of Rafał Sokołowski's profile on X / Twitter

I realized something weird today. I spend more time thinking about what to build than actually building. Feels productive, but it’s just procrastination disguised as strategy. You only learn when you ship.

How to start when you don't know where to start? Pick something small. Do it. Learn. Adjust. Repeat. That's the whole game.

How to instantly build trust and create craving? After my launch, one guy who does security audits reached out to me by email. He pointed out a potential security issue (let's be honest guys, it's hard to cover all edge cases). He explained how he'd fix it (give value) Mentioned that he might send a few more issues he found (create problem) Then asked if I'd be open to working together (follow-up offer) It's definitely a time-consuming approach, but this is how you build trust, educate customers, and offer real help.

You know what's underrated? Building momentum before you have results. You don't need a big win. You just need to keep moving. Tiny steps compound faster than motivation ever will.

Today I crossed the magic 100 followers. It's been a pretty fascinating journey. Sometimes I asked myself if it even makes sense, cause the growth looked so tiny compared to some growth hackers on X. What's the key to success? Honestly, just consistency. I tried to show up every day and it compounded. I believe things will start to accelerate from here!

Do you think AI makes us dumber? I talked recently with a friend of mine. He told me that every Saturday he does a 4-hour block of deep focus work. He asks himself questions he wants to think about, then tries his best to answer them (without using Google, chatGPT, or any external sources). You struggle at first, but after a while you start coming up with answers - or at least interesting insights. It's a way to train your brain to actually think. Want to stay sharp when you're older? Do some blocks without AI, where you just think by yourself.

Time to wrap this 30-day challenge. Of course, I could write a 2xA4 page about everything I learned, but let’s stick to the 3 most important things 1. Ditch the fear of making mistakes I'll be honest - my app had a lot of bugs (probably still does). I built fast and dirty, didn’t overthink my tasks, and just tried to ship as much as I could. That approach was necessary considering the time limits (both the 30 days and the daily time available - not to mention vacations xd). Were there bugs? Yes. Did anyone get hurt because of that? Nope, no one even noticed. Thinking about it this way helped me move with way more confidence and speed. Sure, this advice probably wouldn’t work for a 10-year-old VC-backed startup that runs on reputation, but for someone just starting out and learning - it’s the most effective way to go. I wasn’t mad when someone pointed out my mistakes - I was actually expecting it. 2. Find who you’ll serve and what problems they actually have The project started as a solution to my own problem. I solved it... but during the challenge, I realized I didn’t really need that solution anymore. And being honest with myself - even if I did have that problem again, I probably wouldn’t pay for the solution. It’s just not a “take my money” kind of pain. So yeah - build a painkiller, not a vitamin. But it’s hard to build a painkiller without first figuring out who it’s for. Find your niche, get genuinely curious about it, understand their pain, and only then start solving it. 3. Validate fast You’ve got your niche, you understand their problems, and you’re ready to solve them. But remember - at this stage, it’s all just assumptions, not facts. You might feel like you’re building a painkiller, but the real validation comes when people are willing to pay for it. That’s the info you want to get fast… really fast… Think of it like this - 1-2 hours of research and a week of validation can save you 30 days of work. You hear it all the time - “validate before you build.” I always imagine people who say that as magicians. Everyone says it’s possible. I’ll try that approach (with some of my own twists) and see for myself if it actually works. So yeah, that’s it for this 30-day ride. Built fast, broke stuff, learned a ton. Now it’s time to take all this chaos, add some focus, and keep building. Let’s see where it goes next 🚀

At the beginning of your journey? Treat almost all decisions as reversible. It's unlikely you're facing choices that can't be changed. If you think this way you reduce fear. Make a mistake? Apologize, fix, and move on. That's how real learning happens.

This is true pain no one talks about

and then wonder, why your product doesn't generate revenue

Most people don't relate to your success. They relate to your struggle. The messy middle, the false starts, the doubt, the "what if I quit my job today?" thoughts. That's the story worth sharing.

3 hours of debugging will not omit you 🫡

What would you recommend to someone who doesn’t have a unique angle (yes, let’s assume somebody like that exists) for finding their target audience?

Day 6/30: building a useful X reply-writing partner Man, trying to move the needle while working a 9-5 is tough... Yesterday I focused on polishing the UX with shortcuts. Today I kept stacking small but important pieces: ✅ Sidebar ✅ Reply history UI + backend ✅ Minor fixes to the main workflow (using it daily keeps sparking new ideas) But - hit a snag. Calls to the LLM API feel extremely slow compared to normal chat apps. Since I'm returning a structured object, streaming isn't an option, and I'm not sure yet if the lag is from db overhead or something else. And because UX is the crown jewel of mine, I need to figure this out asap. Tomorrow will be all about hunting down the bottleneck.

Two types of builders 👨‍💻✨