|
Yesterday, we had Fred Wilson (Union Square Ventures) as a guest speaker in a "CTO Course" at Princeton that I'm the preceptor for. During the lecture Fred mentioned that even though Princeton is just a train ride away from NYC, it's the first time he is speaking here: something he hopes to change. In this post, I'm going to follow up on that discussion (summarize the relevant parts of what Fred said), and discuss role of universities in local tech industry. I'll mainly talk about Princeton and NYC, but similar arguments can be applied to other areas/universities as well. I've often thought about the link between university research and startup activity e.g., the Silicon Valley started on land leased from Stanford University, maintained a close link with Stanford, and over the years not only did tech giants emerge from university research (recent examples are Google, VMWare), but the companies then came back full circle to support the university (funding research, hiring graduates). |
Silicon Valley is considered the top "startup hub", with Boston being a not-so-close second (see this post by Paul Graham). It's not a coincidence that these tech hubs emerged around the top Computer Science programs (Stanford and Berkeley in the Bay Area, and MIT in Boston/Cambridge). Looking at the startup landscape, an interesting recent development is the rise of NYC (as pointed out by Fred Wilson and many others). I'm making sweeping generalizations, but I like to divide tech/Internet startups into three categories:
a) Infrastructure Startups: target large enterprises, and/or build the foundation and "pipes" that others then benefit from e.g., Data Domain (cloud backups), Nicira Networks (network virtualization), and Meraki (enterprise wireless). They typically attract the "tech heavy" talent, are usually located in the valley, and are not as visible as services startups in the public eye.
b) Service Startups: target end users and provide some direct service/application e.g., Foursquare (location), Twitter (micro-blogging), and Etsy (marketplace for crafts). Are based both in the valley and at other places, and often struggle with attracting the top tech talent.
c) Hybrid Startups: build the infrastructure as well as the end-user services. Google directly serves end-users, but innovates heavily in infrastructure-support for these services. Amazon is where you buy Christmas gifts, but is a large player in cloud infrastructure as well. Such companies get the best of both worlds (attracting talent, plus public recognition - think Google).
New York's tech industry is largely services/applications/media startups and has evolved without much involvement from the universities in the area (Columbia, NYU, and Princeton). Fred said that initially he really believed in universities playing a major role in startups, but over the years he has noticed that universities (at least in and around NYC) don't matter that much. Their tech transfer is not very efficient, but he believes that their real value is in the (local) talent pool. I'm going to extend that argument a little, and add my two cents:
Local Community: The single biggest challenge for NYC startups is finding good technical talent. Princeton engineering graduates, for example, rarely consider taking jobs in NYC or even working in the city over the summer. They move on to either the Valley or Seattle, partly because during their time at Princeton they never really spend enough time in NYC to feel a connection or are not aware of opportunities available to them here. This needs to change. It will require both an effort from Princeton and from the NYC startup community. As Fred puts it, there is a need to open a dialogue and increase interaction. He is willing to involve USV and other players, and the universities need to reciprocate.
Industry Feeds Research (and vice versa): On the flip side, without having a vibrant local tech industry, the universities in the area are never going to be able to compete with their Silicon Valley (or Boston) counterparts. The line between systems research and industry work is often blurry and it's not clear which problems are best suited for whom. If Mark Zuckerberg proposes the idea of Facebook, researchers are not going to take it seriously, but years after Facebook's success they might find themselves designing systems that Facebook may (or may not) use. Industry, on the other hand, is usually short sighted and is not going to burn money on solving fundamental problems that don't have an immediate impact. It's the age old "useful" vs. "technically hard" debate, and I'm not going to get into it in this post. The important point is that industry and academia, can however, reach an equilibrium over the long run, where both can benefit from a healthy ecosystem. Stanford has that local ecosystem. Princeton -- despite having one of the best Computer Science departments -- doesn't have that kind of a local ecosystem yet (partly because of it's location). Their best bet might be to look at the closest tech hub (NYC) and try to fill the void of technical expertise needed there.
Here are some quick suggestions from the top of my head:
- Train CTOs: NYC's tech industry finds it extremely hard to find people who can operate on the intersection of technology and business. Princeton doesn't have a business school, but has very talented engineering students. It's possible to take this engineering talent and channel it in the direction of entrepreneurship. JP Singh's course aims to introduce students to challenges faced by CTOs, but there is obviously room for further effort (and on a larger scale).
- Share Data: Princeton researchers might be interested in talking to local companies to get a better feel of what engineering problems they face and/or what engineering problems are important, especially when they try to scale their services and hit millions of users. Researchers usually lack access to real data and real companies have tons of it. There is enormous value in working out ways to share data for research projects (e.g., by removing business sensitive information, and anonymizing data sets, etc.)
- Entrepreneurship Culture: Princeton's entrepreneurship club needs to figure out their "next big thing". It might be taking their TigerLaunch competition to the scale of MIT 100K, or teaming up with incubators in NYC, or something else. More importantly, they need to convince the University, the Computer Science Department, and the Engineering School that building a healthy ecosystem of high tech entrepreneurship is a long term investment, and it's only going to make their academic programs stronger. There is always a tension between academic interests and entrepreneurial pursuits, but in the long run it's hard to remain on the cutting edge of teaching and research, without having close ties to the real world, and successful startups.
- Useful vs. Technically Hard: University research has more overlap with infrastructure startups (technically hard) than service/application startups (considered "fluffy" in the research community). This is not a perfect world, and there is a need to make the best of what's locally available. The "fluffy" things of yesterday (Facebook) can become big players and hit very interesting research challenges. They also end up earning top dollar for their products and some of that money can come back to fund research. As more engineers and researchers participate in NYC's startup culture, their ability to do technically challenging things can go up.
- Better Transportation: The train ride between Princeton and NYC takes only ~60 minutes, if you catch an express train. However, the express trains are not very frequent, there is no WiFi on trains, and the switch to the "Dinky" at Princeton Junction is inefficient. Also, the NJTransit rides are expensive ($33 for a return ticket). Making the transportation system cheap and efficient is a critical long term investment that the local government and the University need to make.
There is a lot more that can be said about this topic, but this post is already getting long. I want to thank Fred for coming and giving a great talk, and thank JP for inviting him.
Disclaimer: All opinions presented here are mine and not of anyone else at Princeton or of Fred Wilson. Any mistakes in quoting Fred's talk are also mine.
a) Infrastructure Startups: target large enterprises, and/or build the foundation and "pipes" that others then benefit from e.g., Data Domain (cloud backups), Nicira Networks (network virtualization), and Meraki (enterprise wireless). They typically attract the "tech heavy" talent, are usually located in the valley, and are not as visible as services startups in the public eye.
b) Service Startups: target end users and provide some direct service/application e.g., Foursquare (location), Twitter (micro-blogging), and Etsy (marketplace for crafts). Are based both in the valley and at other places, and often struggle with attracting the top tech talent.
c) Hybrid Startups: build the infrastructure as well as the end-user services. Google directly serves end-users, but innovates heavily in infrastructure-support for these services. Amazon is where you buy Christmas gifts, but is a large player in cloud infrastructure as well. Such companies get the best of both worlds (attracting talent, plus public recognition - think Google).
New York's tech industry is largely services/applications/media startups and has evolved without much involvement from the universities in the area (Columbia, NYU, and Princeton). Fred said that initially he really believed in universities playing a major role in startups, but over the years he has noticed that universities (at least in and around NYC) don't matter that much. Their tech transfer is not very efficient, but he believes that their real value is in the (local) talent pool. I'm going to extend that argument a little, and add my two cents:
Local Community: The single biggest challenge for NYC startups is finding good technical talent. Princeton engineering graduates, for example, rarely consider taking jobs in NYC or even working in the city over the summer. They move on to either the Valley or Seattle, partly because during their time at Princeton they never really spend enough time in NYC to feel a connection or are not aware of opportunities available to them here. This needs to change. It will require both an effort from Princeton and from the NYC startup community. As Fred puts it, there is a need to open a dialogue and increase interaction. He is willing to involve USV and other players, and the universities need to reciprocate.
Industry Feeds Research (and vice versa): On the flip side, without having a vibrant local tech industry, the universities in the area are never going to be able to compete with their Silicon Valley (or Boston) counterparts. The line between systems research and industry work is often blurry and it's not clear which problems are best suited for whom. If Mark Zuckerberg proposes the idea of Facebook, researchers are not going to take it seriously, but years after Facebook's success they might find themselves designing systems that Facebook may (or may not) use. Industry, on the other hand, is usually short sighted and is not going to burn money on solving fundamental problems that don't have an immediate impact. It's the age old "useful" vs. "technically hard" debate, and I'm not going to get into it in this post. The important point is that industry and academia, can however, reach an equilibrium over the long run, where both can benefit from a healthy ecosystem. Stanford has that local ecosystem. Princeton -- despite having one of the best Computer Science departments -- doesn't have that kind of a local ecosystem yet (partly because of it's location). Their best bet might be to look at the closest tech hub (NYC) and try to fill the void of technical expertise needed there.
Here are some quick suggestions from the top of my head:
- Train CTOs: NYC's tech industry finds it extremely hard to find people who can operate on the intersection of technology and business. Princeton doesn't have a business school, but has very talented engineering students. It's possible to take this engineering talent and channel it in the direction of entrepreneurship. JP Singh's course aims to introduce students to challenges faced by CTOs, but there is obviously room for further effort (and on a larger scale).
- Share Data: Princeton researchers might be interested in talking to local companies to get a better feel of what engineering problems they face and/or what engineering problems are important, especially when they try to scale their services and hit millions of users. Researchers usually lack access to real data and real companies have tons of it. There is enormous value in working out ways to share data for research projects (e.g., by removing business sensitive information, and anonymizing data sets, etc.)
- Entrepreneurship Culture: Princeton's entrepreneurship club needs to figure out their "next big thing". It might be taking their TigerLaunch competition to the scale of MIT 100K, or teaming up with incubators in NYC, or something else. More importantly, they need to convince the University, the Computer Science Department, and the Engineering School that building a healthy ecosystem of high tech entrepreneurship is a long term investment, and it's only going to make their academic programs stronger. There is always a tension between academic interests and entrepreneurial pursuits, but in the long run it's hard to remain on the cutting edge of teaching and research, without having close ties to the real world, and successful startups.
- Useful vs. Technically Hard: University research has more overlap with infrastructure startups (technically hard) than service/application startups (considered "fluffy" in the research community). This is not a perfect world, and there is a need to make the best of what's locally available. The "fluffy" things of yesterday (Facebook) can become big players and hit very interesting research challenges. They also end up earning top dollar for their products and some of that money can come back to fund research. As more engineers and researchers participate in NYC's startup culture, their ability to do technically challenging things can go up.
- Better Transportation: The train ride between Princeton and NYC takes only ~60 minutes, if you catch an express train. However, the express trains are not very frequent, there is no WiFi on trains, and the switch to the "Dinky" at Princeton Junction is inefficient. Also, the NJTransit rides are expensive ($33 for a return ticket). Making the transportation system cheap and efficient is a critical long term investment that the local government and the University need to make.
There is a lot more that can be said about this topic, but this post is already getting long. I want to thank Fred for coming and giving a great talk, and thank JP for inviting him.
Disclaimer: All opinions presented here are mine and not of anyone else at Princeton or of Fred Wilson. Any mistakes in quoting Fred's talk are also mine.
