• by Steve Smith , Staff Writer @popeyesm, September 18, 2009

I don't know if I ever met a "personal valet" or a "gentleman's gentleman" but I sure would like to. I don't mean the obsequious "right-hand man" ideal of Waylon Smithers to Old Man Burns on "The Simpsons." I mean Terry Thomas to Jack Lemmon in "How to Murder Your Wife" or Sebastian Cabot to Brian Keith in "Family Affair." Anthony Hopkins in "Remains of the Day," and all that rot.

The mythical valet was intimately acquainted with your habits. He not only knew where you were and what you needed, but he knew what you needed next. He handed you the airline tickets just as you realized you misplaced them. He passed the cognac and newspaper to you as you entered the front door: the former to the perfect temperature and the latter turned to your favorite section. The personal manservant also gently imposed a sense of order, kept his master on the rails, by filtering out unnecessary noise in his life and ensuring a sense of decorum. Well, at least that was the pop culture construction. As I say, I don't know if such a person ever existed, but the fantasy seems potent and persistent.

The notion of a "digital valet" has been in the air for a number of years. The effort to turn the Internet experience into more of a personalized push medium rather than an intent-driven pull platform goes back at least to those abortive "push media" misfires in the late '90s. In recent years I have heard marketers, semantic Internet researchers, and futurists all predict a "brain-like Web" that can create "concierge" services. In this model, our online browsing and shopping behaviors are layered on top of our social graph, connected to our contacts and calendars. In essence, the full integration of the pieces of our digital lives allow the "brain" to anticipate our next needs. I have heard marketers paint a future landscape in which a service recommends your family's next vacation because it already knows when everyone is free, where your friends are going, and what is affordable to you.

The valet concept is re-emerging on the mobile side now as a massive number of start-ups are trying to leverage the cell phone's locator functionality with crowd-sourcing, behavioral tracking and local services. New applications like FourSquare, Buzzd, loopt, Brightkite and many others are location-aware mobile applications and Web sites that I think are on the cutting edge of the concierge concept now. At Buzzd, for instance, the service is aggregating user input around nightspots ("Buzz") both directly from users as well as open sources like Twitter. The "Buzz Meter" shows you where the true hot spots were. CEO Nihal Mehta tells me that after last week's MTV Video Music Awards, he was able to use Buzzd to find where the hot post-show parties were after the award.

This kind of localized aggregation could move to another level of real service in a couple of years. Once Buzzd starts learning its user's habits and preferences, it could apply filters to the world around you. Geo-location hardware can infer experience and behavior, such as whether you have been in this city before, what kinds of artists you have purchased, food you have searched for, or clubs you have frequented.

It is with that background knowledge that a new layer of service can be added to simple geo-location. If the system knows I am a vegetarian, then that local city map of relevant resources should look very different to me than it would to a steak fan. A jazz fan might be notified that an artist he has purchased before is playing that night in this city, but he also shouldn't be besieged with hip-hop info. If the system knows I have been searching or ordering a lot of Chinese food lately, it might know that some variety is in order and suggest alternatives.

The value of geo-location is not in knowing where you are but who you are. You, your interests and behaviors become a kind of filter laid upon the world. That the point at which information becomes a service, where the immediacy and the physical relevance add unassailable value.

What is interesting to me about these emerging models is that they are at the nexus of the conundrum for behavioral tracking technologies and personal privacy and information control. These kinds of service become most valuable as their knowledge of us becomes more intimate. This is the point at which a vendor has to establish genuine trust and a demonstrable exchange of value with the user in order for both to benefit. It seems to me these are services that crystallize that challenge/opportunity for behavioral advertising: BT in extremis. Personally, I suspect that mobile marketing is going to be a platform that doesn't just extend digital models, but actually modifies them, in ways that will instruct Web-based methods on new approaches.