Jorrit

Steve Jobs on how to perfect product design (genius playbook).

Value is what’s left after you remove all friction.

SpaceX continues to be incredible. The engineering feats are absolutely nuts. What I find more incredible is SpaceX as an organization -- can execute structured long-term research and engineering bets without bureaucracy and with high velocity. 99.999% of organizations at this scale cannot decouple structure from bureaucracy; so whatever @elonmusk and @Gwynne_Shotwell have done to build this is magic!

Most of my customers are IT pros who must write many project proposals to keep their service-based businesses running. Many spend much time finding suitable project references for a given customer request. Ex: “Show me which iOS dev projects you did in healthcare.” I’m building a RAG-based mechanism to connect to their CRM (e.g., HubSpot). They can now input a customer request and get relevant reference projects with a click. After finding references, they can tailor them to fit the request perfectly.

Most of my customers are IT pros in software teams. They looove creating technical documentation after each sprint. To make it easier, I'm building a workflow node to load Jira issues and leverage AI for faster documentation.

Yay! I'm curious how getzenai.com will feel with this new OpenAI model under the hood. 🤓 By the way, how cool is it when the application you are building (e.g., getzenai.com) gets better and better with every new model? 😎

15 thoughts on the infrastructure discussion: "Services like Vercel" vs. "build your infrastructure" 1. Most users likely don't care where your product is running. They want your product to provide value for them—everything else is secondary. 2. Services like Vercel give you instant access to infrastructure. With a few clicks, your app is live and accessible via your .com domain. 3. Without Vercel, you need to build the infrastructure yourself. This is an additional workload that demands your time and focus. 4. Building infrastructure is a lot of fun! It's exciting and valuable knowledge, but it has to be learned. 5. If you prefer building your product over infrastructure, you must pay a few bucks monthly for the services. In return, you get immediate access to infrastructure experienced engineers built. 6. You will adapt your product to the chosen infrastructure. This means that relocating your product might require rewriting a lot of code. 7. You should set maximum budgets on Vercel & Co to prevent them from scaling your bank account away. 8. If your users can't access your product because your custom infrastructure crashes, you must fix it asap. 9. If the infrastructure of services like Vercel goes down, they have to fix it asap. 10. Custom infrastructure can be much cheaper if you know how to build and operate it. 11. As long as your product is small, one day of your time might be worth many months of a Vercel subscription. 12. Custom infrastructure is mostly a cost optimization for me. When your product gains a lot of attention and costs for Vercel & Co rise, you can consider learning and building custom infrastructure to cut costs. 13. Besides lower costs, custom infrastructure can also provide flexibility in building. You can and must build everything your way.  14. And if you want to learn? Sure, then open the infrastructure black box and look inside. It's valuable and exciting! 15. Many excellent guides and open-source tools will help you build your infrastructure. Here's an example by @d4m1n x.com/d4m1n/status/1… Read everything?! Wow thx! 🤗

If I try to solve a problem people usually have daily (or on a different schedule), I want to see them using my product daily. I want to ensure that my products create enough value for customers to keep coming back. Does not say that they are happy. For happiness, especially in the early phases of a product, I interview multiple customers weekly. I'm not trying to quantify customer happiness yet. But the first metric that comes to my mind is the “net promoter score.”

As tech founders, we often focus on product features, but how do we truly measure customer happiness? 🤔 What metrics do you prioritize to keep your users engaged? Share your thoughts below! @heathermoran @patio11 @SaaScribe #CustomerExperience #startuplife #buildinpublic

@alexeykiryanov What should you do if products that you perceive as similar, possibly competition, are priced low?

Sharpening sales skills through classifieds I’ve turned selling stuff into a sport. The game? Sell something for a premium price on a classifieds board. After moving, I had to offload a lot, so I started experimenting: 1.Giving it away for free = major headache. 2.Selling cheap = even worse. 3.Pricing high = slow sales but better results. Now, it’s all about writing ads that sell at top prices. My rules: high price, quick text, no haggling, and only pickup or delivery. Result? My copywriting muscle is getting stronger every day.

@manuelcas000 Happened to me. Changed the dev due to lack of experience. :)

It's harder for me to explain to a freelance developer what I want to do and for them to understand, than doing it myself. I'm talking about small things like buttons, designs, user experience on my websites. Doesn't this happen to you? Today, for example, it happened to me again

@AlexejGerst 💯That’s precisely why I stop using many apps: when data entry feels like friction. I saw a „take picture of your food to track calories“-app somewhere. I found it a nice idea to address the data entry issue. (I never used the app)

@marclou Every product has a negative review. :) Take seriously what matters to you. And: taste is subjective

wake up baby, shipfast drama-of-the-month just dropped x.com/robin_faraj/st…

@rileybrown_ai @Replit @cursor_ai SvelteKit web app with Cursor 👌

For those who have tried @Replit Agent and @cursor_ai What web apps can you build easily with just replit agent? And what web apps do you need cursor? I've found Replit Agent seems to be good at certain projects and struggle with others. Just can't quite describe what type Conversation Tweet :🧵

@webdevcody Thanks 😊

Let me give a crash course on emotional intelligence. When you read a post on X, you probably read with a tone of your own internal monologue mixed with how you feel at the time. If you’re feeling stressed and read something that pokes fun at your favorite tech, you read their post in hostile tone, you get defensive, you reply with an angry comment, you look dumb. Is that internal voice yelling at you while you read? Is it in sarcastic tone? Is the tone serious? Is the tone in a joking matter? Am I reading this from the wrong standpoint? What did he mean by that?! If you take the time to understand the context of the post and user, you’ll find ways to parse through the noise and get the main message. You might even learn something new, or maybe even have a laugh. The moment your face gets red after reading a post is the moment you should recognize your emotional intelligence needs some work. Take a breather, social media thrives from the people lacking emotional intelligence reacting emotionally. Never forget that.

@webdevcody Depends on where you are on the curve:

You’re given a large code base. It works, users are happy, client is happy, your team can still ship features decently easily, but the complexity for onboarding is high, a major library used is unmaintained (but open source), the database choice is obviously wrong for the needs of the app which resulted in complex work arounds to implement features, and we can only allow 2 engineers to do full time refactoring. Do you refactor slowly and ship iterative changes? Leave the project as is and just keep shipping? (Ain’t that broke don’t fix it) Quit the project to get hired to another legacy project because the hiring managers at the new company will probably do the old bait and switch on you? I’m convinced all code eventually grows to a point of “if it ain’t broke don’t fix it.”

Are you new to coding and expect AI (e.g., @cursor_ai) to enable you to build products without learning the technologies you are using? Chances are high that this approach will lead you to frustration. AI is not at the level where you describe features, and AI builds a well-functioning product for you. It’s as if we can 3D print significant parts of houses now. If you have no idea about building a house, a 3D printer that can print parts of a house for you won’t help you. — What’s possible with AI coding? You can generate “blocks of code” very well. (functions, smaller components, endpoints, etc.) But you are the person telling AI which blocks of code to generate. You are the person who has to assemble and integrate those cold blocks into meaningful product features. That’s how you can get crazy productive with AI coding. —  But to tell AI which code blocks should be generated, you need to understand the technology you are using.  This means you need to know the concepts' names, what they do, what they are suitable for, and how different concepts work together. — Also, AI makes mistakes. When this happens, you usually have to unblock yourself. Unblocking yourself is not possible when you don’t understand your technologies. — So, how should you start coding with AI? If you are building a Web App: Learn your technologies.  Start with a tutorial for your web technologies to get a basic understanding.  You can use AI to learn the basics faster and then become productive with AI coding.  But you cannot skip the learning. 🤓🤷 #cursor #ai #coding

I think you can get the same done with ChatGPT but for some prompts you’d need to copy paste multiple files around. High friction. With Cursor: - You don’t have to open up your browser 😊 - Referencing multiple files within a prompt (most important feature imho) - Referencing folders - Inline edits across a file, written by AI into your file

somehow I feel like cursor is not as good as just chatgpt.. ? anyone feel the same?

Hey, #buildingpublic 👋 I'm keeping my promise from yesterday. → Learn Cursor by example - AI coding that works youtu.be/hloUjbCIGxQ?si… Here is a video showing how I have worked with Cursor for months. It's my first video. You'll notice that I'm an amateur. But I can code! Feel free to react! If it's cool, I can make more of it! ✌️ #cursor #ai #cursorai

@_nilni Created a tutorial for you: x.com/jposor/status/…

I just downloaded cursor, any good tutorial to watch?

@ericsmith1302 Same here! It sounds paradoxical, but you end up prioritizing like a champion. Every distraction suddenly becomes less intriguing.

I think becoming a father has made me a much more productive Indie Hacker. I have *way* less time, but I find that it forces me to mentally prioritize what I’m going to work on. I end up working more on the most important tasks because I’m giving more thought to it. I also noticed it’s forced me to build more automation tools so I don’t have to be tied to the business 24/7. I def wouldn’t have prioritized them back when I had infinite time but its really helped with mental burnout. No, I’m not saying having kids is the secret to building a SaaS. Just sharing some perspective I wish I had a few years ago