"A startup is a company designed to grow fast" - Paul Graham Different people have different definitions of what a startup is. But there is only one true definition. The defining factor of a startup is that it is designed to grow fast. If it does not grow fast or if it stops growing fast, it stops being a startup and becomes a traditional business. The idea behind growth at all costs is that you should be making the best use of your time and that means building something that helps a lot of people very quickly instead of taking 10 years to build something helps only a few. How fast does the growth need to be for it to be a startup? The ideal growth is over 10x per year. But this high level of growth comes with one major condition. What you are building should not have been built before because you can only grow that fast if the market is untapped. So while the objective of a startup is high growth, the process of a startup is to solve problems that have not been solved. When you're building something that has not been built before, most of your effort and time will be spent in trying to figure out exactly what you need to build. While you can create theories, ask users, do reasearch, the final test of what works will always be putting something in the hands of the users and seeing if they use it. So the smart thing to do is to skip the theories, user studies, research and go direct to releasing features to the users. Because you can never know with complete certainity if users will use a feature until they use it, your best bet is to spend the least amount of time and effort on the feature to get it to the users. If the users use it, you can improve it and make it whole and bug free and if they don't, you can discard it. It would be wasteful if you built a feature with full fledged UI and functionality and then users did not use it making you discard it. This is the concept of the Minimum Viable Product where instead of building a whole pizza only to find out nobody likes the flavor, you build only one slice of multiple flavors and then build the pizza for the slice that users like. The second most important thing when you are building something that has not been built before is that you do not know exactly who your users are going to be. Every startup has to figure out who their users are going to be after they start building the product. The only thing they have in the start is a gut feeling. You can easily confirm this by comparing interviews of startup founders from when they were starting with the ones after they succeeded and you will see that the founders talk about very different types of users or use cases. So don't worry about not knowing who you're building the product for. Just start on a gut feeling that someone may want the product and then you can later make changes to it based on what they actually want. This is the concept of Product Market Fit. Instead of starting building your product with a completely baked idea of your product and its market, you start off with dough and continue shaping it until you find a product with a market. These are the only two things that you need to think about while building a startup. Don't fall into the trap of unimportant things like Agile, Scrum, KPIs, OKRs. Everything other than these two things fall into place automatically. Of course, there is one condition for things to automatically fall into place. You need to hire the people with highest levels of raw intelligence and skepticism that you can find because these are the people you need when you're building something that has not been done before (paulgraham.com/smart.html). It does not matter what you can afford. If you need more money for this, people will give it you. This is one thing you should not compromise on because if you do, nothing else will work. Users will live with the worst customer support if you have something they want and they can live with an app that crashes multiple times a day or an app they have to struggle to make it to work but they will not live with an app or features they don't need. And these hires with the high levels of intelligence and skepticism are the ones who will figure out the app and features that users want. Otherwise, with increasing number of useless (non-PMF) features your users would use the app less and less to do what they should be doing and after a high enough ratio of useless features, your customers would stop using. Founders often confuse users using non-pmf features with users using the app. Like users on your professional social network app playing games that you have added to it (LinkedIn). To hire people with the highest levels of raw intelligence, there is no objective tool. You have to go with your gut feeling again. But this only works when you are one of these people because it takes one to know one. A person with an 80 IQ will not be able to identify a person with a 120 IQ because he cannot see it. In fact a lot of high IQ people cannot even identify other high IQ person. The gut feeling for this is like the taste for good food. You have to develop it. This inability of low IQ people to identify high IQ people is the reason that B players only hire other B players or C players but never A players. This problem is made worse by the fact that most people think they are smarter than they are. But there is a basic way to self-check. Have you done something exceptional? something that no one else would have? This is one of the questions on the YC application. Or do you understand some truth that others don't? This is a famous question by Peter Thiel. When idenfifying high IQ people try to be careful about people who have have been lost to the corporate world. After working for large companies they have learned too many things that are bad for startups but are too difficult to unlearn. After you have solved this problem and hired people with the highest intelligence, the last problem is to retain them. And this may be the toughest of all the problems discussed until now. Because I have only found some of the reasons these kinds of people leave and there may be many more reasons that nobody knows about. But I'll atleast go through the reasons that we know. One reason these people leave is if instead of having them use their intelligence to figure things out, they are told to do things you want them to do. Which is like hiring the best Formula 1 driver and then giving him instructions on how to drive. This is the surest way to lose these people. Another reason these people leave is if they are made to work with low intelligence people because they end up having to explain and defend the most obvious things. One less common reason to leave is that they want to do their own startup, though this may not have a solution other than letting them build their own startup within your startup like Paul Buchheit built Gmail and Adsense inside Google. Another reason is when the potential upside for them for the startup succeeding is not high enough to justify them losing nights and weekends because in a startup losing nights and weekends is a requisite. This is the shortest complete guide on how to do a startup.