Copying Y Combinator – WHY and HOW

[UPDATE]- I’ve written an update 1.5 years later on the original post: “Looking back – 1.5 years since ‘Copying Y Combinator’”

Have you thought about starting a program like Y Combinator in your city? That doing so would not only build a startup ecosystem but would also bring a good financial return? I studied Y Combinator, TechStars, Seedcamp, and many more programs to develop a framework for “Copying Y Combinator”.

(With apologies to Chuck Palahniuk…)

  • The first rule of copying Y Combinator is: Do Not copy Y Combinator.
  • The second rule of copying Y Combinator is: DO NOT COPY Y COMBINATOR.

The key to copying Y Combinator is to figure out how you can be just as good, but in a different way, than Y Combinator.


Background

This last year I’ve been an MBA student at Cambridge University. In order to complete the degree we had to do a substantial piece of research, and I chose to do it on the rise of Y Combinator and similar “seed accelerator” programs. My hypothesis was that a lot of people/organizations are starting seed accelerators without really examining the full scope of innovations they need to think about in order to achieve their goals.

I wanted to take the opportunity to look into why entrepreneurs choose to go into a seed accelerator, why individuals choose to start a seed accelerator, and then propose a framework for designing new programs.

Key results are described in this post, and the full paper is embedded via Scribd (a Y Combinator company) below.

Data

The very first step in examining these programs is to get data; it’s almost all posted somewhere, but isn’t consolidated. Between the data that each program publishes on their website, press on the various programme Demo Days, and Crunchbase, I built a list of virtually every startup funded by every seed accelerator.

Click this link to see the spreadsheet on Google Docs.

The only accelerator where this data isn’t as comprehensive is for Y Combinator; I’ve put in placeholders where known (ie, 8 companies in cohort X but only 5 have launched). That said, it should only be missing a handful of companies at most. And please note that nearly all exit values are purely speculation, though educated speculation based on exits of similar companies.

I also surveyed the founders of companies that have either been funded by accelerators or are looking to be funded by accelerators (Y Combinator and others). Specifically, I wanted to find out what they cared about when choosing a seed accelerator. The results are as follows:

  • 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

These numbers did vary somewhat between different programs and non-funded companies. (For example, the average Y Combinator founder valued Brand/Alumni connections much higher than the average respondent.) But the trends show that entrepreneurs value the elements of programs that give them long-term chances for success: connections to investors, other connections, and product/business support.

Financial Results

Since seed accelerators are still in their early days, it’s too early to make a definitive verdict on their success. But the early data is promising.

Y Combinator and TechStars are two of the oldest seed accelerators, and are the only two to have had substantial exits. The TechStars exits have likely already generated a profit, and there are several companies that may still exit at some point in the future. The Y Combinator company exits have likely already brought Y Combinator to break-even, even after having funded over 100 companies. More impressive is that there are a good number of companies in the portfolio that could reach substantial exits at some point in the future. (And potentially a handful that could reach the vaunted $1billion+ exit.)

Recommendations – How to Copy Y Combinator

The bulk of my paper goes through the elements that are involved in a seed accelerator program. But the fundamental decisions that can define the potential success of a program are simple.

Success derives from the program’s founders and focus; together they must create a distinctive and compelling reason for entrepreneurs to join them.

There will always be entrepreneurs looking for funding; what a seed accelerator should provide is the right match of resources for those entrepreneurs. If the resources that entrepreneurs get by participating aren’t compelling, the program simply won’t get the highest quality applicants, and thus will not achieve maximum success.

This is why there has been little true competition for Y Combinator thus far: they simply have truly compelling resources to offer through PG and the other Y Combinator founders, the YC alumni network, and the combined program network. Until another program can be more compelling than Y Combinator, they will attract the best startups. (See rules 1 & 2 at top about copying Y Combinator.)

The key when constructing a seed accelerator is to look critically (and honestly) about the resources a founder has available; the founders’ experience and the expertise available to the entrepreneurs. Find the focus point that is different from Y Combinator that makes it distinctive and compelling. For example, FbFund REV accelerates companies building applications on Facebook. The new Springboard program in Cambridge (UK) is focused on B2B software applications.

Maybe your expertise is in mobile technology, maybe it’s in medical devices or maybe its in enterprise software. The key is that the founders and the mentors they assemble for a program in that focus area are a distinctive and compelling reason for entrepreneurs to apply and attend. Once the founders and focus are decided, many other decisions fall into place. For example, the program length and funding level will need to be adjusted so that companies can reach a significant development milestone during the program. Just because Y Combinator is three months long doesn’t mean that your program can’t be 9 months long, provided that’s right for the companies involved.

The full paper has far more detail, the point to take away is that the founders and focus must align, and must align to create a programme where an entrepreneur would travel from around the world in order to participate. (Even if there was an accelerator in their own backyard.) The potential to do this in the field of web applications is diminishing quickly.

Final Thanks

I want to say a specific thank you to the program founders that agreed to be interviewed: Paul Graham, David Cohen and Reshma Sohoni. And a huge thank you to the people that commented on my blog posts and Hacker News posts over the summer and took the survey I described above; your feedback was invaluable!


The Documents

Copying Y Combinator

 

Appendix A – List of Seed Accelerators

Click here to view the list of seed accelerators. Only seed accelerator programs are listed; see main paper for details.

 

Appendix B – Example Seed Accelerator financial model

 

Appendix C – List of all companies founded by Seed Accelerators

 

OR

Click here to view the list via Google Docs.

  • http://twitter.com/Sandymac1000 Sandy Mckinnon

    This is an excellent piece of work – Copying Y Combinator – WHY and HOW – http://bit.ly/4D0UHy – just shows the effort vs progress slope for this kind of stuff!
    BUT it is all goes around; comes around – Good luck #seedcamp -ers; See you Wednesday!

  • http://thegongshow.tumblr.com andrewparker

    If you're going to copy YCombinator, then you should really give credit where credit is due: thank YCombinator. I went to the demo days for LaunchBox, TechStars, SeedCamp, fbFund REV and YCombinator in the past year. The only YC clone that even acknowledged that they were a YC clone and, furthermore, thanked YCombinator for their pioneering efforts was Dave McClure at fbFund REV. All the other programs never even mentioned YCombinator at the demo day.

  • http://blog.jedchristiansen.com jedc

    Interesting point.

    I saw a line on Twitter today from someone at Seedcamp: “When I first set it up, I was stupid. Then, briefly, I was a genius. Finally, it had been obvious.” The same assumptions seem to be in play… in hindsight, Y Combinator fills an obvious need.

    But I understand PG has good relationships with the people at fbFund REV… even speaking to them during the program. I don't think that's the case with the others.

  • mauricebv

    Hi Jed,

    great piece of work, thanks. Maybe interesting to make it a living document so every body can see the latest track record of the accelerators.

    Happy to help out ;-)

    Cheers,

    Maurice

  • http://blog.jedchristiansen.com jedc

    Hi, Maurice.

    I do plan on keeping the list of seed accelerators and the list of the companies they've funded live. If I ever get too busy to keep them live I'll definitely ask for help. :)

    Cheers,
    Jed

  • mauricebv

    Hi Jed,

    Plus cool, I will follow your blog to stay up to date. Let me know when you need an extra set of hands ;-)

    Cheers,

    Maurice

  • http://streema.com/ Juan

    Absolutely awesome! Right now I'm working on my thesis, studying the possibility of starting up an accelerator in a developing country, check it out: http://southcombinator.com , although the main info is on the blog: http://southcombinator.wordpress.com and it is mostly in spanish. :S

    Anyway, Jed, I haven't read the whole work, but I would love to hear your opinion about focusing (as you said about FB Apps and B2B software) in a developing country like Argentina. What do you think?

    Good job man! I promise to read all of it and give you a better feedback. :)

    Juan

  • http://blog.jedchristiansen.com jedc

    Well, I would read my paper for a more complete opinion.

    If you're looking to copy Y Combinator in Argentina, you can certainly do it. But it's not terribly compelling if that's all it is. The best Argentina startups will look to go to Y Combinator or other program to maximize their chances of success, so you'll generally have a lower quality of startups. While it might be successful depending on your goals, it won't be as good as it could be.

    Instead, focus on what you can provide that's interesting and unique. What people do you have as advisors? As potential mentors and investors? What can you focus on that would bring startups from all over the world to Argentina for your program? Perhaps there's something about technology usage in developing countries that can only be learned within one, so any company looking to expand will want to attend.

    It's really a matter of what you want from starting southcombinator.

    Again, I hope that helps. But I would read the full paper for more detailed advice. :)

  • http://twitter.com/marc1919 Marc Nathan

    Jed, I am very impressed with both your research and conclusions. Deep business experience, alignment of interests and a diversity of subject matter expertise have been key factors for the mentors coming together as a group to work with startups in a Seed Accelerator. As a member of http://www.capitalfactory.com and having just experienced our first Demo Day, I can tell you that the social aspects of becoming a mentor are equally, if not more important than any potential financial return. The concept of creating a vibrant startup ecosystem and helping young startups grow is a huge psychological benefit to the mentors that will likely translate to the companies that are being assisted.

    I have put together a Google map of all the known 'Seed Combinators' that I'm happy to share with you for your continued research: http://maps.google.com/maps/ms?ie=UTF8&hl=en&t=

    I am also hoping that my proposed SXSWi Panel on the same subject gets chosen for the conference in March: http://panelpicker.sxsw.com/ideas/view/4859

    I look forward to learning more from you as I plan on watch this blog closely.

  • http://blog.jedchristiansen.com jedc

    Thanks, Marc.

    Your Google map is a great resource. I used a slightly more restrictive “filter” as to what I considered a seed accelerator for the purposes of my paper, but it's good to see the wider variety of programs available.

  • Pingback: blog.jedchristiansen.com » Blog Archive » Recap of feedback on “Copying Y Combinator”

  • Chris

    Hi Jed, well done. I wonder if you considered IdeaLabs in your study and even if not, what are your personal thoughts on their slightly different model. You may be interested to know that in Beijing, a new such programme (modelled mostly after IdeaLabs) is being started by Lee Kai-Fu, Google China's ex CEO.

  • http://blog.jedchristiansen.com jedc

    Chris, do you mean IdeaLab, the one founded by Bill Gross? If so, I didn't consider them.

    If you read my paper, you'll see that I specifically addressed the seed accelerator model that Y Combinator pioneered. Y Combinator seeks out, funds, and supports new businesses from across the world, investing money and time for a stake of the business. If you look at IdeaLab's site, they specifically state: “Idealab is not accepting outside business proposals for review.” As it's a completely different model, I didn't address it.

    My personal thoughts on IdeaLab is that it is very difficult to get their model correct, and is largely dependent on the people involved. In fact, I'd say it's very rare that that model is consistently successful. But I might be a bit of a pessimist…

  • Genny

    Hi Jed,

    Great work, a lot of useful info, but I have a question. I don't understand why, at the financial model part, the accelarator has a diminished shares: from 5%, to 0.5%, 2%, 4%… Why is that?

  • http://blog.jedchristiansen.com jedc

    Hi, Genny.

    This is a good question, and has to do with general dilution of ownership stake.

    Accelerators all invest in (typically) the very first round of funding for a particular ownership stake, generally about 5-10%. (YC's median is 6%) What's important is that they do no other funding of the company.

    Successful companies generally go on to have more and more rounds of funding. Since accelerators don't put any additional money in, their ownership stake is reduced with each new funding round. I assumed that as a company becomes more successful they go through more rounds of funding (which isn't always the case), which means the accelerators have the lowest ownership stakes in the most successful companies.

    I hope that makes sense!

  • Genny

    It makes a lot of sense, thanks for clearing that up for me.

    I though that, if the company's value increases, the accelarator's share decreases, which didn't make much sense :) But it's all clear now, thanks again.

  • http://blog.jedchristiansen.com jedc

    Yes, the accelerator's raw percentage ownership stake decreases, but the absolute value of their ownership stake in dollars still rises. (See the column in the same spreadsheet with the value assumptions.)

    Even though the accelerator is getting diluted, the value of the company is rising faster than the decline in ownership stake. So a hit company does well for everyone. :)

  • http://streema.com/ Juan

    Thanks Jed for your answer!

    I've read it all, and I understand your point about focusing on a determined field, I'll need to research more on that.

    Is difficult to understand how, in spite of globalization and the democratization of opportunities we are living, there are still things that depend on location, and that once someone takes the head, is difficult to catch them up.

    Anyway, as I've said before, your work is a great source of information for me. Thanks for open it to everybody!

    Juan

  • http://twitter.com/rarelyapropos Julie Moffitt

    Jed, thank you for posting all of this information! I just came across it, and am wondering if you've gone ahead with a living version of the data separate from the numbers you used for your analysis. If so, where can we access / view that information?

    Thank you!

  • http://blog.jedchristiansen.com jedc

    Hi, Julie.

    The links above are the “living” data. As I hear about new YC (or other) companies launching I've been updating the files. I may have missed some, so if you know of any others please let me know!

  • http://twitter.com/rarelyapropos Julie Moffitt

    Jed, thank you for posting all of this information! I just came across it, and am wondering if you've gone ahead with a living version of the data separate from the numbers you used for your analysis. If so, where can we access / view that information?

    Thank you!

  • http://blog.jedchristiansen.com jedc

    Hi, Julie.

    The links above are the “living” data. As I hear about new YC (or other) companies launching I've been updating the files. I may have missed some, so if you know of any others please let me know!

  • Pingback: Entrepreneurial Education Programs in New York/Invitation to Founder Institute Investor Session | Investment Banking, Research, Due Diligence, Operating Executives, and Recruiting for Private Equity and Venture Capital Funds

  • Pingback: Seed Funding Database | Sean M Everett

  • Pingback: CO-UP : ??? ?? » Y Combinator ????

  • Pingback: How Private Equity and Venture Capital Funds Grow the Value of Portfolio Companies | Investment Banking, Research, Due Diligence, Operating Executives, and Recruiting for Private Equity and Venture Capital Funds

  • Pingback: An Open Letter to David Cohen of TechStars | The Adventures of Chris Schultz

  • Pingback: Encuesta a Emprendedores – Primera Parte « SouthCombinator’s Blog

  • Pingback: Why the seed investment bubble is exactly that at Elias Bizannes / blog

  • Pingback: Yuri Milner: If Yuri Milner had made his investment offer to all previous Y-Combinator companies, how would that investment be doing today? - Quora

  • Pingback: Which incubators exist in Silicon Valley? - Quora

  • Pingback: How Private Equity and Venture Capital Funds Grow the Value of Portfolio Companies | Founders Space

  • Pingback: incubators are a ghetto « fate = will && choice || circumstance

  • Pingback: Startup accelerators in 2011 — blog.jedchristiansen.com

  • Pingback: Looking back – 1.5 years since “Copying Y Combinator” — blog.jedchristiansen.com

  • Pingback: A new home for seed accelerator resources — blog.jedchristiansen.com

  • Pingback: How to start a business accelerator — MBA disseration « ebusinessaccelerator

  • Pingback: Five baseline assumptions on seed accelerators — blog.jedchristiansen.com

  • Pingback: Business Accelerators – Part II « Random thoughts on technology, finance and innovation – kalliagk's blog