In the last few months, I've encountered at least 3 people who are thinking about incubating. On the one hand, it's amazing that the topic still comes up, but on the other hand, it's not so amazing.
Clearly the news about the spectacular failures of incubators during the last Internet bust hasn't deterred anyone from trying to build incubation operations. We can even find examples of incubation-type operations that are arguably working like Ycombinator and similar operations in other locales, like Techstars and LaunchBox Digital.
People keep wondering themselves why they can't just build stuff, and keep building stuff until something works. Or setup something like Ycombinator/Techstars/LaunchBox. It seems so cheap and easy to build a web site/service and get it out there. Why not just build lots of things quickly and try them until something works?
I believe the answer is YES you can. But before you do, I would point you to my two blog posts about incubation, which were gleaned from conversations from many incubator operations both personal and groups in the present and in the past. At betaworks, I took on the project of finding out as much as about incubation as I could, to inform betaworks about the best way to go about coming up an idea (or ideas), how to get them built, and set them up for later success. I then went out and talked to people incubating and distilled what I learned into these two posts:
Words of wisdom/caution to those who want to do this:
1. Generally, professional investors run away from people trying to raise money into incubators, because of lot of them were around when the dot-com incubators all collapsed and they remember that. So calling whatever you're doing an incubator has been found to be a detriment to fund raising.
2. Find your benefactor, woo someone who believes in you and what you're going to do and has cash to fund your operation.
3. Be aware of destroying incentives that will hinder people from working their butt off on their projects. These are things like paying people full salaries and benefits, and not tying the ownership of the project deeply enough to the people working on them, among other things. Keep people hungry and motivated, that they think their entire future and life depend on the project they are working on.
4. Transferrance of an idea is SUPER HARD. Don't think it's easy to just be able to explain an idea to someone else and they will understand it as intuitively and viscerally as the idea originator.
5. Keep costs as low as possible. This will keep everyone's runway as long as possible. It will also add to the hunger.
6. Figure out what you're good at, and leverage that in developing the incubation operation. Paul Graham loves smart hackers and is really good at filtering for that. So Ycombinator teams are always composed of super smart hackers because his strategy is based more on having super smart hackers around than just the idea, because often times where you start is not where you end up, and smart people will find a way to success no matter what. What do you believe will generate success and how will your strengths help that strategy?
7. Following on 6., I would advise you not to compete with Ycombinator or any of the other incubator type operations out there. Don't call yourself Ycombinator 2.0 either in name or messaging. Only Paul Graham can pull off Ycombinator in the way he is doing now. Be yourself and build your own brand.
8. Out of respect for the kind folks who shared with me their knowledge and wisdom which cost them a ton of time and money in legal fees to figure out, I am not publishing anything about what I learned about company structure or legal matters. You should go find a law firm who has worked in this area before and they can help you figure things out. As a hint, some of that has to do with the SEC Investment Company Act of 1940.
I wish you well in your incubative endeavors. May you build something truly great!