In your restaurant and cookie example you have a well defined market. If you're a restaurant you are really a Chinese restaurant, or a Mexican restaurant, or wh

In your restaurant and cookie example you have a well defined market. If you're a restaurant you are really a Chinese restaurant, or a Mexican restaurant, or whatever. You have defined your market by type of food and price point. In a startup all these things are up for grabs. You have much more freedom to experiment, but it is orders of magnitude more expensive to test business models then it is to test cookie recipes. One of these takes months or even years to validate, and one of these takes less than an hour. "The second thing is that I get the feeling that the author simply haven't explored the huge "product space" between takeaway and ready-to-cook prepackaged ingredients with a recipe at all." This is very true, but that ties into my point: it's much much harder to do this testing than to test a new recipe. Hence many people try to find a cheaper way to validate ideas, which is what market research is all about. Now whether the OP did market research well is a different matter. In fact he's quite up-front he didn't. But that doesn't mean well conducted market research in not worthwhile.