Home

Build

I'm going to say something I've never said out loud before. I divide the people in the world into builders and non-builders. And I look down on the non-builders. Anyone who goes out and builds something is immediately raised to gold status. I will go meet builders in their city, I will drop almost everything else to talk to a builder. I see the act of building with a kind of religious reverence. Also because of this, one of the biggest sins in my book is to be insincere about what you're building. By being insincere, I don't mean not working on it full time or prioritizing a job over building something. By being insincere with a product, I mean someone building it for reasons like money, status or investors. The only reason of building should be to build a thing for users. It doesn't have to be thousands of users. You can build something for yourself and I have the same level of respect. In fact, almost everyone starts by building something for themselves. (The second step is to build something for yourself but it is something that some other people also want) Of course there's a third intermediary category of people as well. The ones that I see as potential builders. These are the people that I have very strong hopes for. I hope they will build something but I can never know for sure because I'm not sure what divides the builders and non-builders. Is it a lack of agency or is it their complacency to the problems around them or is it something just temporary like being unable to find a problem they can actually solve. Anyone who's smart should be trying to build things. There should not be a single smart person that does not try to build things. There are a lot of reasons for this. Building products has a very big upside. If you're smart, you want to actually benefit from being smart. I like to think of people that are super smart like Formula One cars. Sure a Formula One car can be used as a towing car and sure it's going to be better than most other cars at towing, but is that the best use of a Formula One car? The best use of a Formular One car is only on a specialized track built specifically for the Formula One cars where they can hit their limits for which they were built. So sure a super smart person can do what everyone else is doing and be better at it than others but is that the best use of that smart person? Like Formula One cars, building products is a specialized track where they can reach maximum utilization of their brains. The second reason super smart people should try to build a product is that building a product has an unlimited upside. There is no other arrangement in the world where your upside is unlimited. In a job, it's limited by the structure of the company and your boss. In a service business, it is limited by the number of hours you have in a day. The third reason to build a product is that you get to solve your own problems and the problems of others and make money while doing it. The money is a nice by product that makes the entire process sustainable. Because you could also build products as a non-profit but it wouldn't be sustainable and you wouldn't be able to continue doing it long term or on the biggest scale. Now moving onto building. Builders are very unique people. Because all the kind of things that make someone build something are exceptional. Because a person needs to first face a problem and then instead of accepting it, think of ways to solve it. And someone who builds a solution for one of their problems has already thought about solving hundreds of problems because only after you've thought about solving hundreds of problems do you find one that you can actually solve yourself because not all of the problems you want to solve will be solvable. A very common thing among builders and potential builders is that at some point they would have thought about solving education and making learning better. It is such a common pattern that there is literally no builder I've talked to who's not thought about trying to fix learning or education. This might go into a tangent now but learning cannot be fixed. The problem is that the education system is an international monopoly of credentials. And it is one of the oldest held monopolies. The problem with trying to build a better solution for learning is that the academic system does not care and people therefore have to prioritize getting the credentials and degrees instead of focusing on actual learning. So even if you made a 100x better alternative of learning, people are still going to go to the same schools and colleges to waste their time in lectures because they need those credentials to join the professional world of jobs. Though I do think there is an indirect way to fix learning by building companies that focus on learning instead of credentials. But this itself is a very tall ask. While it is not uncommon, it is not common enough to break the academic credentials monopoly yet. Coming back to builders being unique, there is a small problem where all the startup advice tells people to build for a niche. But this advice is counter productive because builders are already so unique that what they were going to build was already going to be for a niche anyways and after this advice they end up building for an even smaller niche where the total number of people with that problem in their country is probably less than 100. So for people who already know they're very unique or very weird, try to generalize to a slightly more average user when you start to build something. If you're a kid, you can build a product yourself with Claude. If you've never looked at a single line of code before, it's going to take 2 months for your brain to get the hang of it. Don't try to crash course it in a shorter time. Because the brain works in a specific and rigid way. Code is like maths in that it has a lot of arbitrary symbols that mean something. And it takes some time to get the hang of it like it takes some time to get the hang of the symbols in calculus. You should do this now because the context required to build products has been reduced. You could not have done it before in 2 months without LLMs. Don't try to do courses on SQL or some programming language or bootcamps. They're wasteful distractions. If you're someone who is more senior and has a job and a family and does not have the time to build something himself, you can start by paying someone to build it for you. But please avoid going the route of the software development agencies. You want a kind of person that can build the whole thing himself without any intermediaries. Because if it works in the long term, you would want to bring that person on board full time. So don't work with someone you wouldn't want to work in the long term if the thing works. Dev agencies have too many problems, the biggest one being communication overhead and the second biggest one being that they make the most amount of money when they pay the developer the least amount of money which is obviously not great for either the person paying for development or the developer. The dev agency model works mainly for bigger projects and not startups. You should do this now because it is now possible to have one person build the entire product because LLMs now allow some developers to write more code in less time than a team of 10 engineers could before. And you don't have to try to manage an entire team of developers while working at your day job or have it done through a development agency because of this. Now coming to the actual building. The only correct place to start is to have an idea of a product that you want but doesn't exist in the world. Start by trying to build that, no matter how badly. It will take you 2 months the first time regardless if you're working on it fulltime or part time because if you've never looked at code, that is the minimum period your brain needs to actually start getting it. I've spent a lot of time studying cognitive science but this number isn't some exact calculation. It's a rough estimate. For more details on basics of how learning works, see the book A Mind For Numbers After you've built a product that solves a problem for yourself, you're going to have to talk to more people to figure out if it also solves a problem for them. This is where things get tricky. Because people don't think about what they feel and people don't say what they think. If you tell someone you've built something, they're going to be nice and try it just to keep your heart and tell you it's very useful. But that's a problem. It's like asking your mom on feedback on your product, your mom is obviously going to praise it even if she can't tell what it is. That's why your next step after building the first product should be to figure out how to talk to other people about their problems related to your product so you can figure out if other people actually have that problem and if your product solves it for them. There's a short book that is written specifically to help with this problem of talking to people about their problems and about your product so that you get the actual feedback that can be useful for you. The book is The Mom Test More often than not, you'll realize that the problem people have is very different than the one that you're trying to solve or that your solution doesn't actually solve the problem in a much more meaningful way than other solutions. In that case, you move on to building a better solution or to solving a different problem. Builders also face an inherent problem at this step because builders are the type of people that are obsessive about things. If they've found a problem, their brain doesn't want to let it go without solving it. And this then becomes their Alice in the Wonderland moment where they keep trying to fix the problem harder and harder but don't get anywhere. Another pitfall when building a product for the first time is that because the builder spends a lot of time working on the solution, they lose focus of the problem and they keep trying to make their solution work for the problem. They often even end up changing the problem to fit their solution. This problem is also inherent to building because you will naturally spend more time on the solution because it will require you to work on the details of building. So in your head, the solution takes up more space because it has more parts and more details and the problem ends up taking less space than the solution. And the brain is a dumb machine in this case where it doesn't care what you care about. It cares about the thing that has the biggest footprint in it's memory. And finally how do you decide if you've built something that people really want. The measure of this is not plain retention because you're at a very early stage. Half of your users are going to be people you know and they're going to open the app just to see what's new even if they don't want to use it. The only measure of this is that you need to find those few users for whom your product is a matter of life or death. Users that use your product in a way that they can't live without it. To identify these users on a large scale, there is a format created by the founder of the Superhuman app. They basically ask all their users the following question: How would you feel if you could no longer use Superhuman? A) Very disappointed B) Somewhat disappointed C) Not disappointed Basically, people who reply they would be "Very disappointed" if your app stopped working are the ones that are your target users and those are the ones that you should optimize building features for in the future. And the percentage of people that respond with this option shows you for what ratio of your users is your app actually useful. I have an interesting story about how we did this at Markaz. I picked 100 users which were representative of the types of users in that it included sets of users who were doing low sales, medium sales and exceptional sales. I sent all of them automated WhatsApp messages with the question formatted as above. The weird thing was there was this one woman who was doing exceptional sales worth 100k per month and she responded with "Not Disappointed". Obviously after every response, I would call them and talk to them in detail and it turned out that the user primarily sold branded products and she could do that outside of Markaz as well by buying clothes in bulk in sales and then selling them again to earn a very high margin. So this is was someone who could invest good amount of money to do this kind of work. Our model was especially geared towards allowing users to sell products directly to their customers without buying them first themselves so they did not have to do inventory. An investment free business. For context, Markaz is a social drop shipping app where users(resellers) share photos and details of products on their whatsapp status in Pakistan and then the people in their contacts who see those photos place an order with them. The resellers place orders for their customers address directly by adding a profit margin and Markaz delivers the product directly from the supplier to the customer and collects the payment and then sends the reseller their profit in their easypaisa account. Coming back to this exceptional reseller. While our app was making them good money, their life did not depend on it and they already had alternatives because they could invest money. So obviously this was not our target user and we did not optimize for her. Now the interesting part, there was other reseller who had only started recently and had not made a lot of sales. When I sent him the message, he responded "Very disappointed" and so I called him. And when I asked how his experience with Markaz had been until now, he said sir mainay aik gents suit aik banday ko mungwa kar dia tha but kapra itna rough tha ke customer jo mera jananay wala tha usko use he nahi kar sakta tha. Tou mainay woh suit apne paas mungwa kia socha khud he stitch karwa kar pehan lu ga. Lekin sir woh itna bura kapra tha ke mainay aik dafa pehan kar wash kia tou kapra he khatam hogaya. Multiple jagah se phat gya. After listening to this, anyone would normally assume that the person would never use our platform again. And so I thought maybe he didn't understand the question I had sent on WhatsApp and I tried asking it again, agar Markaz app kal band hojaye ya aap kisi wajah se kal ke baad use na kar sakein tou apko kesa feel hoga, but I hadn't even listed the options yet and the uncle was like "band horahi hai app!!!?", "kyu band horahi hai app!!?", "sir band nahi honi chahiyay main agay use karu ga!!". So I had to spend like 5 minutes just to explain it was a survey question and the app is not actually going to stop working tomorrow. Now you may wonder what happened here. So did I. So I spent a lot more time trying to figure out what was going on. The thing with this user was that for him, the ability to find products and sell them to his peers was something that he could not do without our model because he would have to buy inventory with his own money, money that he didn't have enough to spare, and then he could also never be sure which products would actually sell and if he was unable to sell some of the products that he had bought, it would be a huge loss for him. So for him, our app was a matter of life and death of this business that he wanted to do. So we obviously optimized for him in the long term. So now, closing the discussion about how to identify if you've built something people want and how to identify the people for whom it is a matter of life or death, you can use this Superhuman format. P.S. I tend to unfollow anyone that shares a certificate that they got. Certificates are almost an anti-building trait. Collecting certificates is fine if you want. But if you share a certificate with the world, it means the things that you think are worthy have a very low bar.