3 ms·

Craigslist is, unfortunately, the "best" example of the network problem. Due to its popularity, it is entrenched. That said, there are methods of disruption. A

Craigslist is, unfortunately, the "best" example of the network problem. Due to its popularity, it is entrenched. That said, there are methods of disruption. AirBnB bootstrapped off Craigslist to get their initial dataset, and have since moved to their own content. That's probably the strongest approach if you're lucky enough to be able to move quickly when you need to.

scarmig 14y ago
The thing is, I think Craigslist's network effect is exaggerated. Its network is very clustered because nearly all of it (particularly housing) is focused on people in a particular area. Compare to Facebook, which has a large degree of inter-connectedness between all clusters. If there were a site that had, say, all the rentals available in the SF Bay Area, it would be highly useful to you, as it would have gobbled up most of the benefits of the network effect available (of course, outsiders moving in wouldn't automatically know its name like they do for Craigslist, but that's relatively minor). Once you've reduced your problem to "develop a really strong network in a very localized geographic area," it seems to me like the solution would become in reach, with a lot of hard work. You could even scrape pretty heavily early on, or enter it by hand if somehow that got around the ToS. Maybe start at Berkeley or similar college town, maybe reducing yourself to using physical paper and highly targeted online advertising, and rely on students heavily to get the word out. Make it ridiculously easy for someone to post an ad on your site and have it automatically propagate to CL. Gradually expand to different parts of the Bay Area, and then pick out other localities to target. NYC, maybe, especially if you can somehow make its real estate market less dysfunctional.

3 ms·

Craigslist is, unfortunately, the "best" example of the network problem. Due to its popularity, it is entrenched. That said, there are methods of disruption. A

by dclowd9901 14y ago

scarmig 14y ago
The thing is, I think Craigslist's network effect is exaggerated. Its network is very clustered because nearly all of it (particularly housing) is focused on people in a particular area. Compare to Facebook, which has a large degree of inter-connectedness between all clusters. If there were a site that had, say, all the rentals available in the SF Bay Area, it would be highly useful to you, as it would have gobbled up most of the benefits of the network effect available (of course, outsiders moving in wouldn't automatically know its name like they do for Craigslist, but that's relatively minor). Once you've reduced your problem to "develop a really strong network in a very localized geographic area," it seems to me like the solution would become in reach, with a lot of hard work. You could even scrape pretty heavily early on, or enter it by hand if somehow that got around the ToS. Maybe start at Berkeley or similar college town, maybe reducing yourself to using physical paper and highly targeted online advertising, and rely on students heavily to get the word out. Make it ridiculously easy for someone to post an ad on your site and have it automatically propagate to CL. Gradually expand to different parts of the Bay Area, and then pick out other localities to target. NYC, maybe, especially if you can somehow make its real estate market less dysfunctional.