(This is mostly based on this post from 2000.)
There are two types of startups you can create. You have to know from the start which one it’s going to be. If you can't decide, or start using the rules from the wrong kind of startup, you’ll probably lose.
- Stack Overflow, Facebook, Yelp, Amazon, Trello
- has to grow quickly
- concentrates on getting as many users as possible in order to earn money in the future in some way
- relies on VC money
- new technology, “land grab” - you have to conquer a lot of territory quickly, or a new competitor will show up who will take more and you’ll be left with nothing
- usually relies on network effect - a lot of people have to start using it for it to be useful (because they generate content that other users will want, because they recommend it to friends, etc.)
- often uses a lock-in of some kind, in that the users are less likely to leave if they stay for a while, because of the content and friends they have in the system
- does not start generating profit until it gets really big
- there’s plenty of money, so it solves any problems by throwing more money at them (e.g. by hiring someone who can solve them)
- it can have problems if it hires a lot of new people too quickly because they can get out of control
- has to move quickly, even if it means breaking a lot of things and making a lot of mistakes, has to experiment and learn from those mistakes
- 1% chance to win $10B
- there are a few hundred companies like this
- you want to get VC money, but get only as little as you need to grow in the near future and get it as late as you can (so that you buy it for a smaller share of the company and give less control to the investors)
- there are various types of investors that invest in companies at different stages of development, with a different set of risks and deal with different amounts of money
- investors are evaluating companies based on the idea, the people who work on it, its prospects for development and the possible risks (that it won’t work, no one will use it, no one will pay for it) - remove some of the risks, show them that it can work and earn money, and you will quickly double or triple the evaluation
- if you’re asking everyone and no one wants to invest, go back and work on it some more, remove the risks etc.
- Fog Creek Software (Fogbugz)
- concentrates on earning money from a smaller number of users from the start
- relies on its own income (bootstrapped company)
- a market with established competitors, you’re not trying to take a new territory, but take existing territory from them little by little
- gets break even quickly
- even if there’s just one customer, it’s valuable to that customer
- there’s no money to waste so it tries to save it whenever possible
- tries to be careful, mistakes can kill you because if you lose income you have to fire people
- 90% chance to win $10M - this will probably make you more happy unless you’re a gambler
- there are millions of companies like this - they’re much much more common, but the big ones get more press because they’re cooler and more interesting
- money from this kind of startup can let you build the other kind (like in case of Trello)
Tip for Polish companies: think globally, don’t create products for Polish market only because this puts a limit on how big you can grow.