Where 2.0 preview: Whrrl shows the way

We've covered Whrrl, and several of its competitors, already on Webware, but with the Where 2.0 conference coming up next week, I thought it'd be interesting to dive into this product just a bit more, since it represents some very interesting trends that are central to the creation of location-aware apps.

Whrrl is a fascinating project. The idea is that it tracks where you go, through your mobile phone, and makes that information available to your social network if you allow it. It also uses the behavior of other Whrrl users in general, and your friends in particular, to generate recommendations on places to go, and things to do based on behavior--not data entry.

For example, if you frequent a particular restaurant, you don't have to do anything; Whrrl will know it. But it will also know which of your friends go to the same joint and what other restaurants they hang out at. And then it will be able to recommend just those establishments to you.

It will also have the capability to show you where your friends are in real time. For an old guy like me, that's pretty creepy. But I think a younger generation of Facebook users might find it a natural extension to social networking.

Here are the trends that Whrrl illustrates, that I think we will see echoed a lot at the Where 2.0 show:

Collecting geodata has to be passive

There are interesting services that rely on users entering in their location, but that makes for a very different experience and, more importantly, a horribly incomplete data set.

If you want to get good data, you've got to do deals with the mobile carriers

Only a few handsets, like the BlackBerrys, allow developers access to raw GPS data. For almost all other phones--including the iPhone with its upcoming SDK--getting geo data means going deep into the machinery, and you cannot do that without the carriers or manufacturers giving you access. Sad, but true.

People have social network fatigue

No matter how cool a new mobile social app is, requiring people to sign up their friends to make it work will throttle its growth. A better bet is to allow people to leverage their existing social networks, like Facebook, as a starting point. But even then, new mobile social apps require some sort of buy-in from social net users to work. Until the big social networks like Facebook and LinkedIn get their own geo apps (not add-ons, but actual integrated features), I predict that growth of the mobile social products will remain limited.

Privacy is everything

The geo app companies are going to have to do a crackerjack job of "stewarding" (as Whrrl CEO Jeff Holden says) the personal location data they collect. They're also going to have to educate users on managing their geoprivacy. For an interesting take on this, see FireEagle. (See "You are here, sort of.")

Geolocation means more than GPS

Relying on the global positioning system only gets you so far. For one thing, only a small (but growing) proportion of existing mobile devices are GPS-enabled. For another, GPS is not accurate enough to identify behavior. Whrrl, for example, uses raw data from GPS receivers (on the phones that have them), in addition to proprietary analysis that includes group analysis of other users' signals and "dwell" time at certain coordinates, to determine not just where users are at a given moment but whether they are inside or outside a building.

Other geolocation technologies that matter are Wi-Fi location (see Skyhook Wireless, which Whrrl is partnering with), cell tower-reported location (which only provides a rough position centered around the tower), multiple cell tower triangulation (more accurate), RFID-reported data, HDTV signal-based geolocation, and of course user-reported location. Smart geo apps will be agnostic to the input method of geo data.

I'll be scanning for interesting new geo apps at the Where 2.0 conference next week, and will be sure to report on the Launchpad event Monday night when several companies launch new initiatives.