|Challenges that JFDI faces to its model||Hugh Mason||6/22/14 7:36 AM|
I was asked this question in a private email and thought it better to answer it here in case anyone else wants to chip in with an opinion too:
What challenges has the JFDI accelerator program faced till now?
I guess there have been several phases, each with different challenges. Meng and the rest of the team might want to chip in here:
Meng and I met and came to the conclusion that we were meeting a lot of well-intentioned people at all the business plan competitions we were being asked to judge, but those people didn't seem to be meeting each other. It was Meng's genius to see that, without a physical place where engineers in particular could meet each other, Singapore was not going to get to critical mass. So he and some younger guys got together and set up Hackerspace.sg, which was the island's first ever coworking space during the day and a geek hangout at night. The conversations we had and the fantastic friends we made there were hugely important in giving us the courage to try doing an accelerator. As was David Cohen from Techstars, who made a rare visit outside the US to attend one of the first, if not the first, echelon conference (I recall it was about 100 people in a room somewhere). David encouraged us to learn from what he was doing and give it a go.
So we incorporated on 10 JAN 2010 I think. The first big challenge was to secure funding to get going at a time when nobody had done an accelerator anywhere in APAC. We were aiming to run a techstars-style roughly 100 day program so we set ourselves the challenge to close at least one major funder within 100 days, or else we'd say that the market wasn't ready, we weren't the right team or whatever. We had a lot of help from the community, not least through this online forum, and we got a written commitment from Singapore's Media Development Authority on day 94 as I recall.
Although Lean Startup and its terminology was just a twinkle in Eric Ries' eye at that time I guess you could say we identified our MVP as being to test the mentoring component of the accelerator, which is still what really adds value. We wanted to test if we could match world-class mentors with great local teams and to create enough engagement to make that worthwhile for both sides. It turned out we could, so while we kept on fundraising, we moved to test the next critical element.
A key concern was to test if there was a sufficiently large and ready-enough pool of talent in the region to fuel an accelerator. It was clear from the outset that Singapore's small population would make businesses and a trawl for talent just on the island unsustainable - JFDI had to be a regional play. So what was out there? There was no data, research and no blogs covering the whole region at that time. Just local people meeting up for beer and peanuts in lots of cities. The geek grapevine proved invaluable in connecting us to them across the region as we went to barcamps and the like and made a lot of friends.
And we were also helped by David Cohen's invitation to become one of the first members of what was then called the 'Techstars Network' (but is now the Global Accelerator Network), which aimed to find a scalable way to share what Techstars had learned because David was getting drowned in requests to meet and talk about it. We could see that Techstars' pioneering work was beginning to settle down into a method: it wasn't clockwork but there were some processes and insights that seemed to create repeatable results and reduce risks. But would it work outside a few special cities?
It was easy to believe that someone could make one of these new-fangled accelerator things work in Silicon Valley (as Paul Graham had done with Y-combinator) but we were more interested in learning from Techstars because cities like Boulder, CO where Techstars took root were, on the face of it, much less unique. It was clear that a lot depended on the local ecosystem - in Meng's phrase, a lot of the secret of Silicon Valley's success was for 'inventors to become investors and founders to become funders', sharing the spoils of success rapidly with the next generation. Would that happen in Asia too, or would filial piety mean everyone who achieved an exit stashed the cash in a family trust fund, interested only in backing shopping malls and oil palm plantations?
So during this year the major challenge was to pull together enough cash to fund operations because the startup ecosystem in Singapore was a lot less developed at the time. Investors were rightly cynical about an unproven, complicated model - they could see that it had a hell of a lot more moving parts than just bunging some startups in a shitty old space and throwing in pizza once a week with a speaker. That model (traditional incubation) was manifestly failing. So it was very important to get people like the TenCube and JobCentral founders, who had exited locally, on board with us and we remain eternally grateful that they chipped in because they believed in it.
But it was still unclear if we would find enough talented teams. It's hard to remember now but in 2011 many countries across APAC had not even held a Startup Weekend, let alone seen their first coworking spaces emerge. With sponsorship from SingTel Innov8 and local partners in every city, we helped to run six Startup Weekend events in Singapore, Jakarta, Manila, Gurgaon (just outside Delhi), Bangkok and Melbourne. That convinced us that there was loads of energy out there in the community and a burning desire for people to come together around the emerging coworking spaces and meetups. And it also revealed a harsh truth: that just because someone goes to a Startup Weekend doesn't mean they necessarily want to be an entrepreneur, just as going on an art weekend doesn't mean you want to be a professional artist or going to night school to pick up some flower-arranging technique doesn't mean you want to be a florist. Startup Weekends, hackathons and the like are fantastic for introducing people to each other but they were never designed to nurture and feed accelerators.
So we came to the conclusion that there was enough talent to run an accelerator but also that we'd need to figure out how to feed it into the future.
In January we had scraped together enough cash and talent to run our first full accelerator batch and running it was a riot. I documented my experience through it in a Video Log that captures the countless operational and other challenges that we faced, day by day. It was a lot of fun and it worked - more than 60% of the teams went on to raise an average USD550k+ within six months of the program ending.
It left us exhausted, of energy and cash, so we took the second half of the year off to raise more cash and work out what had just happened and what we could learn from it. The biggest challenge then was to convince ourselves and everyone else that it wasn't just a one-hit wonder and that we did have something repeatable. We realised that it's one thing to cobble together the resources to do year 1 but following through consistently is much harder, especially as the competition for talent around us started heating up.
By the end of 2012 the gospel of Lean Startup was spreading fast and everyone wanted to start their own business, and their own accelerator. Every worn-out incubator and academic teaching program rebranded itself as 'an accelerator' and all kinds of dubious operators set up around the world - landlords with crappy spaces they couldn't rent, exploitative fund managers looking to rip off gullible young people, corporations with crazy ideas that the world would come flocking to them if they put their brand on a dodgy scheme with no substance. And some countries round the world started shelling out startup capital like candy, without any of the support structure it needs to do its work wisely. I think we are still going through that period of hype two years later and the crunch is coming. Many of our challengers locally and internationally are failing and that the collateral damage is not going to be pretty.
As we kicked off two more batches of startups in 2013, it was clear that if we wanted to make a go of this for the long haul we would need to prepare for the fall-out and make sure we were offering something robust in every sense. We focused first on trying to bring as much rigour as we could into our selection and acceleration. We started swapping metrics regularly with other GAN accelerators and I finished up a Masters course which I used to tie what we were doing into all the academic literature I could find that was relevant, in an effort to try and make the thing scientific wherever possible. That has been a success and we can now say with much more rigour why the stuff that we and our mentors had been preaching based on empirical experience now works.
Our team discussed our three main challenges for 2014/2015 in a week-long series of internal workshops and Meng and I were delighted that everyone wanted to push to overcome them. The challenges are to:
So from day to day the challenges we have faced so far this year have been the classic ones of any startup making the transition to be ready for growth. We aren't all wearing suits yet (except Meng) but we do now have a lot of standardized processes and systems for example. We're also now held accountable, rightly so, by our investors and in particular IIPL who have been incredibly supportive but who also need us to deliver results for the whole ecosystem.
What changes do you foresee to adapt the accelerator model to SEA needs?
A key realization has been to understand what we are. When Meng and I set up JFDI we thought we were running a process - like a carwash for entrepreneurs. They would come in one end, mentors would shampoo them, rinse and repeat, then they would come out shiny for demo day. As I have written in comments on one of our key mentors' blogs we now realize that actually an accelerator is about community management. It's about bringing together people who want to share the risks and rewards of innovation, using science to minimize the risks as they take lots of little bets of time and money to probe the unknown.
Looked at that way, it's clear that an accelerator is a social construct, so we are still working out how to make that work in Asia's many different cultures. Even things like Lean Startup don't always work right out of the box here. For example, in some of the countries of SEA, asking people about their problems in the way you would in a US-style customer discovery interview would be seen as incredibly rude. The underlying methodology is right, we think, but the way you apply it has to be different.
And of course the community of investors and mentors is different in SEA too. For example, in countries like India it's still very common for traditional investors to expect to own a majority of a startup from the seed round onwards. In countries like Malaysia friends tell us that there is less of a tradition of unpaid mentorship. And across the region fear of failure is a barrier to innovation. All these things can probably be overcome but it does mean that the evangelical folk from the US and Europe who frequently show up on our shores tend to find that their miracle programs end up a flash in the pan. We think it's about working bottom-up, with what is, rather than trying to introduce something alien.
Has $25K funding, 3-months program, 5-20% equity worked well?
In some ways yes, in other ways no and the key one is reflected in your question - you didn't mention all the intangibles, like the mentoring and the community, which is where our true value lies!
Nobody should come to JFDI for $25k. That's just to make sure a team can eat and not live on the street. But in today's price-comparison style of shopping, it tends to be what everyone focuses on. So startups in cash-rich countries like Singapore might see us as poor value compared to a government grant and delude themselves that because they have been given a stack of cash they have somehow been validated (they haven't - look at the failure rate of companies that get early funding that way). We still have a messaging problem there. And the converse is true in some of the poorer countries around the region where $25k is almost enough to retire on. We have had some teams come to us feeling that they have succeeded simply by being admitted ... that's a challenge we still have to overcome in a very diverse region.
As for the 3 months duration, we are looking at being flexible on that now that we have back-to-back programs running, our own dedicated space and the opportunity for teams to take time to do the customer discovery through our JFDI Discovery program, taking however long that needs to get confidence before they push the pedal to the floor with the JFDI Accelerate. It helps that we now have a much more thorough understanding of the steps that every startup has to go through to succeed: we may not be in a position ever to 'pick winners' but we do know how to identify the flaws that are very likely to sink a team. Those essential steps are reflected to some extent in the valuation table we publish to be as open as possible with teams about what the local market thinks they are worth on the basis of the evidence they can show now.
|Re: [OpenFrog] Challenges that JFDI faces to its model||Wong Meng Weng||6/22/14 10:17 AM|
On Sun, Jun 22, 2014 at 10:36 PM, Hugh Mason <hu...@jfdi.asia> wrote:
Scale Out can be broken out along longitudinal and geographic dimensions.
Geographically, we want to serve more of Southeast Asia. Our founders come from Indonesia, Thailand, Philippines, Vietnam, India, Malaysia, etc, etc, etc. Also, France. That's fine if the startups are global and define their market in terms of language. But it's challenging when the startups are serving local markets – it's a bit of a disconnect to do that from Singapore. Yes, they should be able to sell online. But Lean Startup talking-to-customers requires more than getting out of the building – it requires getting out of the country!
Longitudinally, we see founders who aren't quite ready for JFDI Accelerate. So we launched a pre-accelerator program: JFDI Discovery. At the other end, we see some startups which have already received seed funding. Their next milestone is Series A, but they still need help to get there. To make things work at that level, we need to invest $100,000 or more per startup. Other accelerators have the funds to do that. We don't, yet.
Existentially, a final question is: of the big success stories of recent years, how many are attributable to an accelerator, and how many were indie startups which did it their way? That's the billion-dollar question – the "emperor's new clothes".