Here’s some irony: By going national, AmericanTowns has become more local. The company, founded in 1999 by former Time Warner exec Mike Kelly (who later went back and became president of AOL Media Networks) was one of the original hyperlocal sites.
Kelly’s idea was to build out a template, serve affluent suburban communities, and have local reps — typically housewives or teachers — selling ads door to door. The site soon added local search to the mix via a partnership with Planet Discover.
Like other local sites before it (i.e., Citysearch), the company soon grew frustrated at the lack of scale and slow growth of advertising. It stripped down some of its uniquely local characteristics to build out a national model (something CitySquares has done more recently). That model was appealing enough to get $3 million from Idearc in spring 2008 for a minority share (and perhaps more importantly, rights to LocalSearch.com, which was owned by AT&T).
The cynic in me said the company would wind up subsisting on Google revenues, and would barely be more local than a typical geodomain site. Fifteen months later, Chairman Ted Buerger tells me it is doing far, far better than that.
Buerger says the company has gotten real traction over the past four months, gaining 50 percent more users, and now gets more than 1 million unique visitors, per Google Analytics. It is also getting 2,500 announcements per week from community organizations, and has more than 40,000 events submitted each month from community leaders who are plugged into the site.
The site’s additional usage, adds Buerger, was achieved by aggregating 1,500 blogs, and the aforementioned partnership with Idearc’s Superpages, and de-emphasizing its ghost towns. More importantly, it refocused on its most successful towns, like Pleasantville, N.Y., 35 miles from New York City, where site logs indicate it has 60 percent household penetration on a monthly basis, and regular usage by 25 percent — about the same as a typical newspaper, if taken at face value.
“More successful towns grow more rapidly,” says Buerger. And then they spread around a circle via word of mouth. Nearby Ossining, for instance — the home of Sing Sing prison and short story writer John Cheever — clearly saw the success of Pleasantville. It approached AmericanTowns and now has its library, school board, and town and village government up on a shared AmericanTowns community calendar.
Successful towns also tend to have better usage of the weekly events e-mail that gets sent out. To be sure, the AmericanTowns e-mail for my town of Carlsbad, CA, is a boring recitation of weekly Rotary and church group meetings that would make me want to move (if that was really all there was). But Buerger says that there are a number of towns that really utilize the e-mail feature. “Posts automatically appear in the e-mail and people share information,” he notes.
Along with increased usage has been a boomlet in advertising. Buerger says the site gets high local RPM rates in the $16 to $17 range, and national rates ranging from $8 to $11. The primary advertisers are local — as intended — but there aren’t local sales forces. Most sales are handled by agencies, or are self provisioned on the site. The site is also beginning to sell advertising in the e-mail newsletter, having recently sold a package for Realtors.
Buerger’s theory is that a lot of the raised consciousness around the site, and other local sites, has been caused by the increased use of mobile local media services. “It is definitely accelerating things in local,” he says. “We saw a [mobile] campaign by Toyota in the Southeast. It really struck us. It demonstrates that advertisers want to get at the local level.” Buerger says he expects to get more traction in June, when the site comes out with a redesign.
To keep things in perspective, AmericanTowns is still run by a handful of people — the same number of people it had a few years ago when it initially de-emphasized local. We don’t really know how much advertising is being sold, and it isn’t clear that the site is out of the woods yet. But if it can claim Pleasantville-type numbers in other communities, the site may still prove out the original model, updated for local 2009.