kevin p. siu

For the past summer, I’ve been working at the University of Toronto, with Professor Parham Aarabi and his startup ModiFace, which does facial recognition and related technologies. It’s been an enjoyable summer, and I’ve learned far more in this short period than I have in the past two academic years at the university. Although I’m not sure if that says more about the school or about the work I’m doing. The next couple of articles will be devoted to what I’ve seen over the summer. This particular one contains some ponderings about my views of startups.

It’s an unavoidable fact that in my wanderings with the computer world, I’ve come to hear a lot more about startup companies than the average Joe. After all, nearly the entire computer industry is based on a few highly successful startups. Thus, I’ve been ever fascinated about the startup world, the supposedly wonderful world of lawlessness (in the no-bureaucracy sort of way). Admittedly, I didn’t actually work at one of those fancy Silicon Valley jobs with the extravagant parties and CEOs in T-Shirts, but I did catch a glimpse into a similar sort of organization. These are the startups that have the business types all scrambling, trying to figure out what makes them so great. What business school and case studies will never teach you, of course, is what it’s actually like to be in a startup.

Actually, I never intended to work at a startup. I signed on as a “Research Assistant”, that most venerable of all job titles in the academic world – it means you’re the monkey at the bottom of the chain pushing buttons so that the suits at the top become rich and famous.

Therein lay my first lesson about startups. Professor Aarabi, as smart and hard working as he is (more on that later), still had far more things he wanted to do than he could do. Evidently, at these small companies of just a handful of employees, you are always short on manpower, but big on ideas. For Jeff (my computer neighbour) and I, there was always a new project to work on, no matter how quickly we did our work (and we definitely became highly proficient at our work, especially on the web programming front).

The constant need for more manpower can never really be underestimated. To be successful – certainly in a startup – you have to be willing to put in the work. The old adage that you need to work hard to succeed is never truer anywhere else. For someone in charge of such a small company, it is nigh impossible to delegate to others, since the ideas, the visions, and the skills are all exclusively theirs. It was not uncommon for Professor Aarabi to whip out an idea out of the blue and have it fully implemented in code within hours, even if those hours were (for us) overnight.

Startups need to be nimble. You never know what your customers will want next, and you never really know who your next customers will be. Whatever the case, you need to be quick on your feet to survive in a startup world. There are millions of startups, especially in this “Web 2.0″ technological age. There will be competitors, and you always gotta stay ahead by moving quickly.

I got so used to this quick pace of work that, when I took a step back and reflected on my previous experiences, and the experiences of some of my friends at some rather larger banking-type corporations, I was surprised to see how much I’d accomplished in such a short time. I was astonished by the lack of efficiency and expediency of these large moving dinosaur companies. How on earth do they survive at all?

What’s great at a startup is that there are really hardly any rules. There’s no chain of command to report to, there’s no proposals and reports to write, there’s no big evaluation of teams, there’s no use of big clumsy software, and there’s definitely none of all that bureaucratic nonsense in a big company. This point has probably been beaten to death by the esteemed writers in all the tech journals I read, but it should not be lost on the business-y types that efficiency is extraordinary when there’s no rules to follow.

It was great, because I could choose my own tools to write my software, use my own styles and conventions, and leverage any experience I’d had plus any ideas I could come up with quickly. This made things a lot quicker (for me), and a lot less stressful. The whole experience became satisfying very quickly, even if the work I was doing was not what I’d envisioned.

ModiFace’s apparent success had me wondering for a while. I wasn’t really sure how successful such apparently simple technology could be. (Well, simple for the users, but there’d been a lot of programming to make it so simple) If you’re starting a business, it turns out the most important thing is not how good your idea is – plenty of people have great ideas every day that don’t amount to much – but it’s how you can sell your idea, and who you can sell your stuff to.

ModiFace had found their niche pretty well. Facial detection technology, while groundbreaking and by no means an easy problem, is not the most attractive of all the crazy emerging technologies of today’s world. It hardly makes any noise in the science and tech world, and to investors, it’s not immediately clear where revenue could come from. The genius of all this is not to market to an already jaded bunch of nerds/geeks who have seen all the Web 2.0 startups, but is to go for the solid, middle-to-old-aged-stay-at-home types – a demographic virtually untapped by the Web 2.0 “revolution”.

Another strategy also well employed by startups nowadays is to market to other small companies. Not as glamorous in terms of fame, but definitely bigger in the fortune department. For example, IBM stopped selling computers to consumers long ago, but they make an f’n killing on selling to businesses. Startup 37signals also doesn’t give a damn about consumers – they’re a virtual unknown outside the tech business industry, but they also make millions easily through other startups rather than individuals.

Given this much thought (there’s a lot more in my head where it came from!), I almost want to find some great idea and venture into that startupland myself. Getting a good idea… easier said than done though.