Y Combinator will accept 10,000 startups to prove there’s nothing magical about Silicon Valley

The first YC batch, August 16, 2005
The first YC batch, August 16, 2005
Image: YC
We may earn a commission from links on this page.

Y Combinator was a heretic among venture capitalists when it backed its first eight companies in 2005. Co-founder Paul Graham later wrote that his philosophy was that “investors should be making more, smaller investments, they should be funding hackers instead of suits, and they should be willing to fund younger founders.”

Heresies piled up from there. Graham, along with his wife and co-founder Jessica Livingston, accepted applications from anyone, not just via personal introductions, as almost all venture capitalists do today. They chose inexperienced college students over startup veterans. The “batches” of startups were given small investments and expected to work wherever they could find space near Harvard University, where YC was born.

Y Combinator, which has since relocated to California, was virtually ignored at the time. Today, its approach is practically conventional wisdom. The accelerator-turned-venture fund values ranks among the most successful in Silicon Valley. Its mission to mint new startups is copied by hundreds of accelerators around the world. YC says its portfolio is worth about $80 billion 1 with more than 50 companies worth at least $100 million and several above $1 billion, such as Airbnb, Dropbox, and Stripe.

But ask YC president Sam Altman if the program has achieved its vision, and the answer is no. That will require one more apostasy: start thousands of new startups without ever coming to Silicon Valley. Altman says he’s trying to prove it’s possible later this year by accepting 10,000 new companies into its new online Startup School.

“The original, long-standing goal of YC is that startups are good for the world and individual, and how can we get more of them?” said Altman in an interview in early May.

Today, the assumption is that Silicon Valley enjoys an alchemy of talent, ambition, money, and mentality few can replicate (though not for lack of trying). This Silicon Valley terroir, this theory goes, is as unique to startups as Napa’s wine country up the road is for great vintages. Countries and cities struggling to recreate Silicon Valley shouldn’t waste their time. And neither should accelerators that don’t offer funding, exclusivity, or the potential for serendipitous encounters with the powerful and influential players of the Valley.

Another school of thought sees this as nonsense. In this theory, we’re accustomed to thinking about startups and entrepreneurialism as magic, but we’re better off seeing them as engineering. Raw talent and startup potential are distributed everywhere in the world. Finding the right way to support them can unlock Silicon Valley-style success.

Naval Ravikant, co-founder of AngelList, a platform matching investors with early-stage startups, subscribes to this second view. As technology upends venture capital as it has other industries, he thinks this will become clear.

“Venture capitalists like to say there’s only a few good new companies every year. I looked at my physics textbook but couldn’t find that law of nature,” he said. “The idea that innovation is limited to small number of people who gather in Silicon Valley is ludicrous. … The number of startups is only limited by human ingenuity and the market’s ability to absorb them.”

Scaling the Y Combinator way 

For now, YC can only turn out about 300 companies per year through its bi-annual program in Mountain View, California. In technology terms, YC itself has not really scaled. Most of what it does remains analog and manual: mentoring in classrooms, weekly dinners, and a program requiring everyone to live in California for months at a time.  

So a few years ago, Altman and Graham began plotting ways to expand YC. One of the first attempts was a series of 20 lectures on starting companies, taught by Altman and guest speakers at Stanford in 2014. That was followed by an 8-week fellowship program in 2016 (now discontinued) offering a light-weight version of YC with $12,000 grants and access to advice from the YC community.  

Although none were breakout hits, the experiments were promising. In April 2017, StartupSchool.org was born. The 10-week massively open online course (MOOC), billed as a free online version of YC, promised to ”teach people about how to start a startup, and equip them with the needed resources.” The first cohort was made up of 3,000 companies from 141 countries. Participants are not offered YC’s standard deal of $120,000 for a 7% ownership stake, but are given a chance to “replicate much of the YC experience” through a virtual program with mentors, collaboration with peers, and video lectures, as well as receive discounts on services worth thousands of dollars from Microsoft, Amazon, Google, and other companies. YC also encourages the startups to apply to its core program, where they would get financial backing as well.

A screenshot from Y Combinator’s software showing active companies in Startup School.
A screenshot from Y Combinator’s software showing active companies in Startup School.
Image: Y Combinator

Now, Startup School will accept about 10,000 companies in early 2018, says Altman, and potentially more after that. The companies use custom-built software to collaborate, pore over their weekly metrics, and attend weekly mentoring sessions with YC alumni (200 volunteered in the first cohort to meet with participants). Lectures by executives and employees from WhatsApp, Stripe, Reddit, Facebook, Khosla Ventures, and Sequoia Capital are taped and broadcast through Stanford’s CS 183F class ”Startup School: The First 100 Days.” Mentors hold group sessions with founders on Google Hangouts to hash out problems. 

YC claims Startup School has cracked the code on a new type of educational institution to start companies (only without “runaway costs and declining educational standards” of universities, says Altman). It claims only 20% of participants have dropped out after eight weeks (companies must complete 90% of weekly updates and office hours to stay in the program) compared to a typical attrition rate in MOOCS of 87% (pdf).

“I think we’ve stumbled on this free, lightweight online version of YC,” says Altman. “We’ve really been able to replicate a significant fraction of the YC experience.” 

Sandiep Medisetti, founder of alumni relations startup Almabase in the first cohort, said half the value comes from interacting with peers at their own companies over video and text chat. His mentor, a YC graduate who volunteers several hours per week, has been instrumental in solving customer-relations challenges for Almabase’s 100 or so clients. “A couple of my friends in India are doing startups who are just going clueless and do not know what to do,” he said. 

If Startup School works, YC wants to grow its model to global proportions. Theoretically, there’s no inherent limit to how many companies can join YC if its core value is collaboration and online advice (the ratio of mentors to startups at the moment is 30:1, although YC plans to increase this).

For Avichal Garg, a former director of product management at Facebook, the not-so-secret sauce is YC’s methodology. ”There’s a bunch of things that YC got right which are still true,” he said after a Startup School lecture on a sunny afternoon on Stanford’s campus this May. Garg said the handful of principles that YC shares freely online are simple but proven to work—from identifying a clear problem to constant experimentation and reporting weekly results.

“Lots of what they are doing is teaching the same thing to people over and over again. That’s the really interesting bit about this,” he said, referencing countless blog posts, videos and slide decks. “The content you can find on the internet. But putting the structure around it to deliver it and create the right incentive structures around it—you get that right and it can scale.” The difference, he argues, is like reading an exercise book versus hiring a personal trainer. 

Give the whole world a startup 

Accelerators supply young companies with education, mentorship, and financing for a few months to speed up their growth. But is YC (or any accelerator) the right way to mass-produce more companies? Researchers who study accelerator programs have a tentative answer: it depends.

The best accelerators do work, says Susan Cohen, a management professor at the University of Richmond in Virginia, but buyer beware. Cohen tracks accelerator performance at Seed Rankings. She found that top firms generally deliver on the promise of speeding up a company’s fundraising and customer growth, but there is a wide range and the lowest performing accelerators can even harm a young company’s growth. Simply adding more zeros behind the number of companies in accelerators will not work. 

Cohen calls YC’s Startup School “a phenomenal experiment,” but she believes to succeed YC will need to sustain traits she’s seen in all great accelerators. “The best are able to help even the most confident entrepreneurs understand that, while they have a very promising idea, they still have quite a ways to go to fulfill their vision,” she said. “Then, back up that motivation with resources for the entrepreneur to learn what they need to achieve their goals.” If it does fail, she suggested, founders will catch on soon enough. Since young companies need talent, funding and customers to grow, something Silicon Valley still offers like nowhere else, programs must ensure companies don’t fall off a cliff after graduating. “The market is efficient,” says Cohen. “There’s a sorting: successful [accelerator] programs will survive, and the other programs may not.”

At the heart of YC’s philosophy is the premise that the world has too few people solving fixable problems, a “bottleneck in society,” says Garg. “If you fundamentally believe that, then [Startup School] is exactly the right thing to do. You are empowering a bunch of people to just try and it’ll turn out it’s going to be way easier than they thought it was going to be. And there’ll be more progress.”