I came away from TechLeb with some very interesting mixed messages about incubators, science parks, technology trade zones, whatever. (Jacques Masboungi’s talk on the subject was particularly interesting.) On the one hand, they seem to be one of the best things governments can do to foster technology development. On the other hand, they seem to be one of the easiest ways governments can screw up. And since no two projects are the same, it can be hard to generalize from experience.
Given all that, I shall now proceed to theorize about how to construct an environment for fostering technology development. And please note that government does not have to play a leading role. Instead, universities or even private entrepreneurs can and quite possibly should take the lead.
1. Technology incubators need first-class infrastructure. That’s a requirement, not a competitive advantage. They’re competing with developed-world facilities. In particular, bandwidth is utterly crucial — so much so that, if the country generally lags in that regard, “preferential” access to and pricing of bandwidth need to be arranged.
2. Similarly, technology incubators need to have or have proximity to first-class amenities. The individuals you most want to have working in them could, if they chose, probably get jobs in the US. And they’ll need to have meetings with individuals who visit from the US/Europe/etc. Gymnasiums, fast food, etc. — whatever techies expect in Boston or Frankfurt should also be available in the incubator. Indeed, this may be even more important than in developed countries, because working from home may not be a real option, making the office environment even more important. If you need a checklist of requirements, just take a look around Burlington or 1 Kendall Square or Silicon Valley or the Market St. area or the Flatiron District.
I first realized this point when I learned about how the Indian software services industry has been growing.
3. They should be physically structured so that people keep encountering each other. That’s a prejudice of mine from my university days. I selected my grad school (from among Harvard, MIT, and Chicago) based in part on its physical layout and people flows, and went from there to the Kennedy School, whose new building was carefully laid out for this purpose. I possibly learned more in dining halls and common areas and faculty dining rooms at Harvard than I did in classrooms. (In fact, I’m sure I did, given that I didn’t really go to a lot of classes there.) And the Sundeck Restaurant serves a similar purpose at the granddaddy of them all, 3000 Sand Hill Road. (Which incidentally is a complex owned by Stanford University.)
4. Rapid critical mass is important. Here’s where universities, government agencies, and/or larger companies are crucial — they can commit to being anchor tenants. The secondary services and serendiptious encounters that make incubators work rely on there being serious businesses or other relevant enterprises there in the first place. Somebody has to get the ball rolling. That’s the consistent message I carried away from TechLeb.
5. If business barriers are a problem elsewhere in the country, they need to be alleviated in the incubator. Just cutting taxes for companies in the zone can have a huge benefit — and incidentally can create a critical mass really quickly. The same would go for regulatory issues of all kinds, and for any infrastructure failings. If you are short on bandwidth, electricity, transportation, etc. — well, surely you have enough resources to supply them in a limited area. So pour everything you can muster into one place, and watch it thrive.
The classic model for an incubator is a “public/private partnership,” often with a university as a major partner. I suggest that while serious capital commitment needs to come from somewhere, and some government help is usually needed, government does NOT need to be the driving force, except perhaps when barrier-removal is a crucial aspect. In many countries, an entrepreneur who buys the right land (preferably near a university), hires the right people, has the right government and university relationships, and hires the right tech-business-savvy staff could make a huge difference on his own. (He doesn’t need tech industry relationships going in; sales and listening ability will do just fine as substitutes for preexisting know-who.)
And here’s what would be best of all — a global investment entity that seeks out local partners and creates such incubators/technology parks, one after the other, in a broad range of developing countries. They would really have the clout to kick-start things with tenants from global tech companies …