Of course, demographic targeting as it's been implemented by search engines (MSN introduced this feature years ago) has yet to really deliver on its promise, because the quality of the data has always been suspect. MSN's demo data has been drawn from subscription services such as Hotmail, which likely contains a fair amount of false information. And while many marketers have likely played with MSN's demographic tools, the fact that MSN is so far behind Google in scale limits the tool's effectiveness.
Google, of course, has the scale, so it's going to be much more interesting experimenting with demo targeting there, especially because MySpace is now in Google's contextual network. So far, only a limited set of sites in the contextual networks are participating, and I have to wonder whether the comScore data used by Google to establish its demographics is 100% reliable. For example, I set up a test campaign that would just target senior citizens over 65 years of age with an annual income of $75,000 or more. Some sites suggested by the tool (such as SeniorFriendFinder.com and IBDB.com, a site dealing with Broadway shows) seemed very appropriate to this audience segment, while others (xBoxCheats.com? HollywoodTeenMovies.com?), seemed to be a major stretch. Obviously, marketers will want to take a close look at sites recommended by Google's tool to ensure that they feel comfortable with the selections it makes before laying any money on the table.
Of course, the real prize will be when Google begins to tap into MySpace's huge repository of user-disclosed demographic data, which has the promise of offering a tantalizingly accurate picture of its membership. While one cannot rule out the prospect that a significant share of My Space's users fake details about their identities, I'd guess that most of the data here is good data. Microsoft, which inked a deal to supply such services to Facebook, will likely not be far behind, and working to direct its demo tools against Facebook's user database.
One large but unanswered question about all this micro-targeting is how the MySpace and Facebook membership will react to targeted ads that to many may seem "creepy." Some will certainly take steps to remove details from their profiles to thwart demographic targeting, but I'd imagine this will be a small minority.
Demographic targeting isn't a panacea, of course, and in fact it may provide search marketers with more than their share of headaches as their teams come to grips with layering this kind of targeting on top of large and already complex keyword-based campaigns using other targeting methodologies such as geo and daypart. But for those who get it right, demo targeting may provide both increased ROI and greater insight into who is actually consuming and acting on their marketing messages. I plan to be following this new discipline with interest and will continue to experiment to see what kind of gains marketers can yield from it in the year ahead, even if it just proves to be a useful tool and not any kind of "Holy Grail."