Copying Y combinator is pointless.

For one to 'copy' it you need the people that make up Y combinator, and even if you could you'd just be diluting the soup, not contributing to it

Don't copy, innovate!

Besides that, the world is a lot larger than just silicon valley, if you can put together a team with 'star attractor' power for would-be founders then you may have a possibility. But without HN (which really is just the top of the feeder of the funnel leading to YC startups) you're going to have a hard time to get the same exposure and mindset in the founders of the start-ups that approach you.

And HN is also a pretty good way to measure the quality of the people approaching you. You're going to have to find alternative solutions to those problems too.

Without a quick 'triage' stage it is very hard to run a seed fund, by making an end run around the process YC has a distinct advantage here. Basically the HN ecosystem is what provides YC with both the start-ups and a quick way to judge the ones that are probably run by level headed people.

edit: funny moderation on this reply, up/down/up/down come on moderators, if you don't agree with it that's fine with me but do me a favour and point out why you don't agree with me.

I did read the article, I think it makes a good point but I really don't think 'copying' a concept that relies uniquely on a few key people is a winning model. Even in principle.

Do your own thing, and do it better. So that way you don't need to think about copying.

> Do your own thing, and do it better. So that way you don't need to think about copying.

This is the exact point I'm trying to make in the paper. Accelerator founders need to look hard at themselves, be innovative, and create their own niche. Trying to just copy YC will simply lead to a sub-standard clone.

Then don't call it copying.

Anyway, your article is fine, it's just that it seems there is a non-existent audience. Those that are dead set on trying to copy YC will do so anyway and those that were busy on different tracks will find that it does not apply to them.

It's like everything else, reuse the bits that you think are good, drop the rest and try to find good alternatives.

What would be a lot more interesting - at least to me - is how YC stacks up scorewise with the rest of the incubators out there. I haven't seen much proof yet that they do structurally better in the long term.

There is a 34 page thesis deliberately answering the question "How to Copy YC" (making the conclusion that he makes in the article).

There is a good lot of that in there.

In the Spreadsheet linked from the article there is listings of lots of funds and their startups. It makes interesting reading; much of it subjective.

For example YC has funded astronomically MORE startups than many other funds. On the other hand very few have exited or obtained more funding - and a large number are completely dead (compared to other funds).

Techstars has a much much lower "dead" rate and a much higher follow on funding rate.

Seedcamp have the "best" success rate (in terms of follow on funding) but fund one of the fewest number of startups.

All of that data though is pretty useless for actually comparing them. For example the Seedcamp follow on funding might be as a result of their manouvering of VC's towards the startups and not much to do with the viability of the product. Or the startups coming through YC might well have a tendancy to fail regularly - but the community of startup operators it creates might be huge; and the connections they all make might ensure their second or third venture DOE succeed with a nice fat exit.

What Im trying to say is comparing the "success" of an incubator is probably an unbelievably tough task.

YC probably has a comparable success rate in terms of follow-on funding, but it's incredibly difficult to find the information for small angel rounds. (TechStars and Seedcamp helped me make sure my data was accurate, which helped their stats.)

But thank you for the comments; putting some hard data and thinking behind this phenemenom was the whole point of my project! :)

Suppose each accelerator funds the top n startups, where n is the amount of slots they have open. In that case, comparing average results would be unfair. Compared to an accelerator that funds n startups, an accelerator that funds a smaller number n' of startups would fund the top n' startups rather than the median n' startups.

If you take the top n' startups funded by YC, where n' is the number of startups funded by Techstars or Seedcamp, how do they compare then?

The problem with this is that you don't know before you accept them what the top n startups will be. Thus, it's a significant decision when choosing how many startups to accept. More startups means a better chance at a big payoff, but with less time devoted to each.

What's important from the point of view of a seed accelerator program founder (versus a startup founder), is what they will know before they accept companies into the program. The "top n" won't be immediately obvious.

I'm somewhat trivializing the selection process, but it's probably more fair to compare the top n startups of YCombinator with the top n startups of TechStars rather than the total program aggregates. One of the things we're evaluating these funds on is their very ability to pick the top n--so a larger fund is at a disadvantage if you compare their median n against a competitor's top n.

I'm excited to hear the conversation around this.

I would like to take a step back and point out that almost every major city in the states has a business incubator. We don't hear about them because not much is produced that we all use. Locally, they can produce some great projects.

Just wanted to say this because the model of small investment, even if it is office space and time, is happening all around the world. YC just did it better than any of them (and deserves a ton of credit). :)

I always go back to the first bubble and study the incubators. Such a sad ending to many of them (and why most 'accelerators' stay far away from the word 'incubator.'

That would be a fascinating read. The take away from that (so I think) would be 'let's focus on building smart and sustainable accelerators' over 'which one is the best?'

Loved reading your thesis, well done!

Thanks. :)

The only thing I found when I looked into incubators is that the business model and thus the incentives for both the startup and the incubator side appear to be much different than in a seed accelerator. That, and the word "incubator" means different things to different people.

My worry is that in some cases incubators can start to become dependent on having companies in their space, and thus there's a lack of encouragement (or active dis-encouragement) to "graduate" from the incubator. That is a significantly different model than a seed accelerator where they are effectively kicked out at a certain date, so both sides have significant incentives to develop a product quickly.

While I haven't studied incubators as in-depth as seed accelerators, the different incentive structure is what I would worry about. Accelerators seem to match incentives better between founders and the program.

I agree that there is still a lot of scope for building more accelerators. They just need to be focused on the specific resources available to the program, and not another YC-clone general-web-application accelerator. Any new accelerator program should aim to be the absolute best in the niche they choose.

The idea that anyone would setup a "clone" without wanting to be better is kind of strange.

No, that's pretty basic.

"It works there and is a success, if I copy it bit by bit I'll get success without trying"

Also, it oftentimes works: studivz.de blatantly ripped off Facebook (to the point where they didn't even change the name of the stylesheet files) for the German market in 2006 and was hugely successful (for exactly the same reasons Facebook was in the US). They were sold for EUR 100 mln to a publisher.

Eventually, obviously, you'll have to innovate your own way, if for no other reason than that someone else copies you and does it.

A big reason why YC is cloned is to create an ecosystem in a different city. (ie, the YC for Kentucky, the YC for Seattle, etc.) They have tended to want to be the YC for that particular city/region.

They then tend to compare themselves to other local startup options (which are probably pretty bad) and not to the full spectrum of startup options, which is the only way they can get the best startups world-wide.

> if you don't agree with it that's fine with me but do me a favour and point out why

Ok :)

I downvoted for this:

I did read the article

Because the point you made is the exact point the article makes - or as I read it anyway :)

The article title is linkbait, the article itself could be summarized in exactly one sentence, and you didn't need to read the article to be able to do that.

In dutch we call that 'kicking in an open door', I don't know what the english equivalent is.

And thanks for coming out :)

Maybe MBA candidates and "seed accelerator" founders aren't so different after all!

Replace the top founder reasons below:

  Connections to future capital: 8.51
  Brand/Alumni connections: 7.83
  Business support: 7.42
  Product support: 7.13
  Pre-money valuation: 5.25
  Level of funding: 4.14

with the analogous reasons to go (or not go) to business school?

  Connections to finance/consulting interviews: 8.51
  Branding from prestigious degree / alum network: 7.83
  Learning and honing crucial soft skills: 7.42
  Technical Learning in the classroom: 7.13
  Opportunity cost: 5.25
  Level of tuition: 4.14

Disclaimer: I have never gone to bschool.

So I'm the OP and author of the paper. I figured I might get criticized for it because it was for my MBA, and there are a ton of stereotypes around MBA's. (Many for good reason, unfortunately.)

But I have my degree in aerospace engineering, have designed and built solar-powered road cars, and used to operate a nuclear reactor. So don't think I don't have my fair share of geek cred.

I understand the sarcasm, but would personally appreciate honest feedback on the content.

Disclaimer: I have gone to bschool.

I don't think he was being sarcastic.

Perhaps not... I'm probably a little sensitive to MBA stereotypes.

As long as you wear your MBA badge front and center you will probably continue to be so.

Hot tip: Your education level is a good indicator of what you are capable of, but not the only indicator and probably not even the best indicator.

So, if instead of trying to 'leverage the title' and looking at yourself as an 'MBA', you'd change that into 'serious professional that is out to make a difference' your sensitivity will melt away and you'll realize that those that use their MBA as a means to differentiate themselves are exactly where the problem lies.

I could say the same thing about my self, "I'm sensitive to people being negative about high school dropouts", because I'm a high school drop-out. Instead, I don't care. You can judge me by who I am, what I've done and how well I've done it. Some of my buddies have PhDs, all kinds of fancy degrees, I wouldn't trade with them for any money (or title) in the world.

Be good at what you do, be excellent if you can and if you're not try harder. And don't let your MBA hold you back ;)

Yikes, definitely not being sarcastic here.

It is an honest comparison of two transforming functions (bschool vs. seed accelerator); the preconditions and postconditions are surprisingly comparable even though the algorithms are quite different.

Yeah... sorry about that. I misunderstood what you were saying.

In that case, the only issue I have is the first transformation. Instead of "connection to capital" -> "connection to finance/consulting interviews", I would instead say "connection to capital" -> "connection to more lucrative jobs". Though perhaps my opinion is colored by the fact that the consulting and finance interviews were hugely cut in the last year, and so a lot of my classmates have gotten interviews/jobs in other sectors. And those jobs are still quite lucrative.

outcome: copying doesn't work, innovate and iterate and you'll be better.

that rule applies to pretty much everything in a startups life.

The only difference is that it takes much longer to iterate a program like YC... on the order of months instead of weeks. It means that more thinking is necessary before you start.

Compared to most other possible ventures an iteration time of months is peanuts. Some industries have iteration times of decades!