I'm Tak Lo. I'm a venture partner at Mind Fund
, a HK-based venture fund. And I'm a recovering startup accelerator operator.
I previously was Director of Techstars
, one of the largest and best-known accelerator programs. With over 20 programs across the world and over 650 companies funded, Techstars is recognized as the crème de la crème
for entrepreneurs. My mandate during my time there was to help set up the London program. At the end of my tenure, we had visited over 40 companies, hired 30 people, and spun out over 4 international programs.
I witnessed the rise of accelerators in London, from SeedCamp
, the Difference Engine and Techstars, to at last count over 24 accelerators in December 2014
. According to Seed-DB
there are now over 235 accelerator programs worldwide with over $12.9 billion USD invested in 5,674 companies. Many claim that the model is broken, but to fully understand this argument, we have to look back to the model’s conception.
There's a certain haziness around when the actual "accelerator" first started. It's a bit like the urban myth of the boogeyman, where every child knows the story but doesn’t know exactly when or how it started. Well, it is generally accepted that the concept began with well-known programmer Paul Graham in 2005 in Boston. Paul selected students with entrepreneurial qualities for his summer program to help them fulfil their potential. He provided angel investment, mentored and guided them, and helped them fundraise, all during the 3 months summer break. It was duly named after a computer science concept - Y Combinator
Meanwhile, an entrepreneur named David Cohen saw this program and tried to convince Paul to try it in David’s hometown, Boulder, Colorado. No dice, was the response. So like any good entrepreneur, David did it himself. He deconstructed Y Combinator to its core methodologies of 4 program elements: a 3 month time constraint, small angel investment for 10 companies, a mentoring phase, and fundraising. His key innovation was to crowdsource mentoring as a way to introduce other ecosystem players into the program and scale out mentoring advice. Techstars was born. And that was the start of the mentor-driven accelerator model.
The accelerator movement boomed from there. David Cohen essentially open-sourced the Techstars playbook and mentor-driven accelerators started operating in other startup ecosystems in the US, and around the world. It was relatively easy for accelerator operators to raise a $3-4mn fund. The mentor-driven model was an efficient way for angel investors to batch their investment activity in a short period of time. Venture firms used accelerators to track the progress of startups before actually investing. Mentors enjoyed dipping their toes in many startups. Startups found the application for accelerator funding fast, straightforward and transparent. This was a win-win-win-win-win for startup ecosystem players. Accelerators were having their time in the sun. But like all great successes, the good times could not last forever.
Dave McClure from 500 Startups took exception to Altman in his blog
. Jason Calacanis, a prominent Silicon Valley VC, had no problem with “incubator hopping” in his blog
. Heaps of analysis
was done to factually disprove Sam Altman’s point. So, why the revolt? Well, it's analogous to a normal distribution curve. There's a top population that's great, a large chunky average, and a terrible bottom. And in this distribution curve, it's generally accepted that Y Combinator and Techstars are the top. Simply put, this advice didn't bode well for the rest. Some even predicted the end of accelerators
. Ultimately however, this kerfuffle is a puff of smoke. Accelerators are here to stay. Like the concept of a democracy as a form of government, the accelerator model attracts a ton of criticism, but there is no better mechanism to build an early stage ecosystem. However, the model itself will morph into a few different areas:
- More accelerators in emerging markets: Mentor-driven accelerators are already saturated in developed ecosystems. But accelerators are still the best mechanism to build early ecosystems, therefore emerging markets will naturally catch up.
- A market correction: Many predict the winter for startups is coming. Some are drawing parallels with the meteorite that destroyed the dinosaurs, but I'm much more level-headed. It will merely be a harsh winter; the better startups and accelerators will survive, and the weaker ones will die. That's all.
- Slight changes in the model: The trend for taking recent college grads is already starting, as seen at EF. Companies like 500 startups' Distro Dojo, are taking companies who've recently graduated from traditional accelerators. The volume of these different models will only increase. Other marginal model improvements include scaling out the number of companies, like Mass Challenge or Startup Chile.
- Rise of corporate accelerators: Many large corporates are in need of innovation and are interested in the startup space, but they have no idea how to even start. The accelerator model, which makes short, time-based experiments possible, allows corporates to engage in a low-risk, low-cost way. This enables continuous learning and provides more information to decide what innovation strategy to pursue.
Because corporations understand that the latest technology has brought massive change to its people, disrupting legacy processes and old technologies in its wake. Without devoting just simple resources, a corporation will not be able to learn how it needs to fundamentally change.