Monday, February 26, 2007

Nielsen GPS-enabled media tracking system measures eyeballs outdoors

I know I've read articles about this type of service before, but Branding Unbound has a nice writeup of Nielsen's latest attempt to measure outdoor media consumption using GPS-enabled tracking devices. Here's the summary of how it works:

[The company] recruited 750 volunteers who agreed to carry the device, called the Npod, or Nielsen Personal Outdoor Device, everywhere they went for 10 days.

Participants, prescreened to establish demographic profiles, allowed their every movement to be tracked and recorded – like an automated digital version of those Nielsen diaries of old. Every 20 seconds, the Npod captured each user’s latitude and longitude, while a computer system compared the data with the coordinates of 12,000 “geo-coded” outdoor signs in Chicago, including bus shelters, standard posters, billboards and overhead signs.

To establish the user’s likely exposure to an outdoor advertisement, the system applies a mind-boggling number of variables – including driver speed, the angle of the display on the road, its distance from the curb, the distance from which the display is first visible, the height of the display, whether it’s illuminated or obstructed, and so on – as users enter “impact zones,” or the point of impression.

Nielsen knows that ¾ of the people who travel under a 200-square-foot sign above a highway overpass actually look at it, while only 30% of those who drive past a bus shelter actually see it.
Using that and other similar insights, Nielsen can then use the collected data to calculate more traditional reach and frequency numbers, which can be used for planning purposes. Given how quickly the out-of-home advertising market is growing, it's not surprising to see these kinds of advanced methodologies being deployed to determine the true impact of outdoor media. To me, the holy grail (and a frightening prospect at the same time) would be combining the impression data for specific users with geolocation data for the specific items being advertised (if they're physical goods, of course). So, for example, how many people drive past the billboard for the local IHOP, and then wind up turning off of the exit to frequent the place for a short stack and cup o' joe?

Another interesting question is how this system will ever be made to work with electronic billboards, which are starting to become more common (though they still account for only a very tiny fraction of all billboard and out-of-home media sites). With a static billboard, you know exactly what's being displayed at all times. With an electronic billboard, you'd have to cross-reference the device's playback log to determine which ad was being shown at the time the individual was looking at it.

You know, because the basic system wasn't complex enough.

Tags: billboards, out-of-home advertising, media measurement

No comments: