What's the next step in evolution of startup incubator/accelerators like 500Startups/YCombinator/Launchbox/etc.? Please use examples from your own experiences and observations.

Given the dramatically lower cost in starting companies, it'd be logical to think that startup accelerators focused on <$50k investment levels might proliferate and evolve, maybe even becoming the dominant new way to start a new company and break into the Bay Area startup community. Where does this go from here? What are the main problems that are left to solve?

In my knowledge, Launchbox shut down its accelerator and turned into a seed-stage VC firm. So I suppose they evolved away from the model, either because they lost primary leadership or because their portfolio companies weren't churning in expected results or both.

Accelerators like YC and 500Startups are trying to "scale" their model and PG claims that they won't stop until they've hit their limits. I've already heard first hand how the recent class don't get as much time with YC partners. But it's all in the network and branding. What I fear most is the accelerator brand becoming more of an "ivy-league credential" and valuation going gangbusters purely because of the vetting process startups are put through rather than the real value of their product.

I disagree with Cheryl that YC is having a hard time scaling. I was in the Summer 2011 batch with 63 companies and I think they did a great job scaling. I have never had an issue where I was like darn I wish I could have spoken to PG/Harj/Garry/etc about this. If you need the help, you can find it.

I'm clearly biased (as a YC part time partner), but have to agree with Stammy. "Time with partners" is based on a queue system, and more time / office hours are created the more requests are made. Also, I've been through YC as a founder three times, including the first class, and it has always been the case that you basically get as much access to partners as you want, which often varies greatly depending on the founder.

It is important to note that it has never been the case that the partners will hold your hand or make you meet with them to "check in" -- generally you are responsible for using YC as a resource as you wish. Enterprising and successful founders make use of this resource.

I'm glad to hear that. I really respect and admire YC and wasn't implying that it can't scale. And I'm certainly not saying that the quality of startups now is any less stellar. It was more a genuine curiosity as to the extent of its scalability, which I assume would have to be largely supported by alumni participation. I do agree with Justin that the onus is on the founders to be resourceful, which is one of the most important qualities YC looks for anyway.

Justin, Paul, it sounds like part of what you're saying is that the next step in the evolution of these incubators is that they might keep scaling up while finding more efficient ways to maintain the quality.

Would it be ridiculous to think of a 500-member startup batch, the size of an incoming freshman class at a university?

Getting into an esteemed accelerator will give startups credibility, resources and access to an amazing network of really smart people, so it's definitely a booster. However, the quality among accelerators differ significantly and some have better models. Ultimately, the journey is a long long road and while getting the right product for the right market at the right time has a lot to do with it, I personally think that founder character & track record is the best indicator of eventual success.

I think it's quite a bit too early to try to figure out what the next step is, given how fast YC continues to grow.

What if YC decides to (e.g.) open another office on the east coast? What if they do 4 batches per year? How much of the value is in the individuals, and how much is in the format and structure?

I think that a lot of these are still open questions, and until we start to see YC itself hit a wall, it's hard to say what'll come next, because maybe the next step is just a bigger YC.

Andrew, I don't think that's a ridiculous idea. In some sense, that's basically what my college and work networks are right now -- there are thousands of Xoogler and Stanford alums starting companies. And, as snotty as it sounds, I'm probably more likely to help them out if someone random from these networks than not if they pinged me. I think the true value is in the network. So, I don't see why an accelerator couldn't build the same kind of camaraderie and massive brand.

As Eric mentioned, I think the biggest opportunity is for incubators to either move beyond the Valley or bring in more and more folks from abroad and send them back after the program. There's so much additional talent outside the SV area not yet tapped by competing accelerators or companies like Google, Facebook, Zynga, and Twitter.

The best accelerators provide significant credibility/leverage and help companies (and first-time founders) get to a clear-eyed understanding and subsequently disproportionate impact for the few things that matter at that stage - customer development, rapidly shipping product, getting early press and fundraising.

As long as accelerators can continue to provide high-quality help and introductions in these areas for their companies even as they themselves grow bigger, they'll be in good shape.

So far I've seen no indication of scaling issues at YC - they're very self-aware and are always a few steps ahead. It almost seems as if YC treats their process as a product, fixing bottlenecks and issues rapidly as they initially start getting noticed. I like it :)