These days, the “consumer in control” has pretty much become the catch phrase de jour in online circles. But what do we mean when we say this?
It would be hard to claim that consumers really are in control -- after all, they don’t always get to choose whether or not they see advertisements or what ads they do see. Most of the time I think “consumer in control” probably means something different: that brands and marketers are working to provide a set of advertising options that are the most amenable and relevant to consumer preferences.
This is especially clear in the search space -- and in fact, text ads fairly exemplify both relevance and choice.
At least, this is the current model. But the search space is heating up, with even Yahoo! jumping into the talks over AOL’s web portal -- and there are bound to be some changes to search advertising in the near future. Text-ads have been so far focused on relevancy -- typing in “jaguar” and getting an ad for a car dealership, for example -- but we’re always trying to make online advertising more user-specific. Search engines may start to do more than just respond to user inputs.
This ad? You'll love it.
There are some indications of this already. MSN has released its own ad-serving technology that combines search terms with age, gender and location information. Earlier this year Yahoo! announced that its Impulse ad-serving program would begin to hold onto search data for 48 hours instead of the previous one hour, and AOL has just rolled out a similar program for its own search properties.
What all three of these programs have in common is the attempt to increase the relevance of their in-network ads by combining a consumer’s search terms with other consumer data. MSN is using demographic and geographic data, whereas the other two are focusing on past behavioral information, but I do think the impetus is the same. And it makes sense: When the focus is on targeting consumer preferences, the more data you have at your disposal, the better your chances of serving a relevant “personalized” ad.
What exactly is personalized marketing, though -- and how, if at all, does it fit into the concept of consumer choice? Take the example of MSN. When a user enters a search term, the ads that show up aren’t just based on that term; they’re partially based on the particular person entering the search. Since MSN can draw on the demographic data it has from Hotmail registrations and other sources, the company can create a profile for each searcher and choose ads better suited for that searcher down the line. In other words, MSFT can personalize the advertisement.
In the earlier search model, ads were based only on immediate inputs: the user chooses the input, and the inputs choose the ads. With personalization, user choice becomes a little fuzzier, since past behavior or demographic data also play a role. So while to choose personalization is to retain some elements of consumer control, it is to bet -- to some degree -- against direct choice.
This might not be a bad wager. Amazon.com works in a similar fashion, and I’ve never heard anyone complain about Amazon’s “recommendations” for future purchases. And Amazon’s A9 search engine has been quietly personalizing search results since its launch last year.
Behavioral or demographically targeted ads benefit both consumers and marketers by making ads more relevant and driving clickthrough rates. But I was curious about the second model as described here by Google. Could retaining -- or increasing -- direct user choice be a hidden boon for search engines?
To answer this question, I turned to a couple of newer search engines who’ve taken the “user choice” model to a higher level than most of their competitors. These two -- Grokker and Vivisimo’s Clusty -- have built their businesses around what is called “clustering” technology. Clustering, simply put, is when web users have the option to categorize search results on the fly. It’s sort of like searching within results, except without the normal time and effort that takes.
Raul Valdez-Peres, Vivisimo’s CEO, told me that not only could clustering improve the user experience, but that it allowed for more accurate ad serving. “The best personalization is done by persons, by themselves, at that moment,” Valdez-Peres said. If a user is searching for a very specific item, their clustering will allow Clusty to serve them a very specific ad.
According to R.J. Pittman, CEO of Groxis, Grokker’s parent company, there are two problems with personalization based on past user inputs. One, user preferences change, and “having a static set of data and personalizing based on that data is not going to add value to the results.” Moreover, Pittman argues that consumers are more comfortable with “dynamic personalization” than they are with the data collection and targeting used by many companies. He says he is hopeful that eventually, “the ads stop being an interruption and become part of the content.”
Grokker and Clusty are taking a different gamble, guessing that by giving users a greater degree of choice they can target ads more closely -- and by increasing consumer comfort level, cause users to respond more frequently to their respective text links.
Different models, same search
What we can take away is that there are two competing models at play here -- the one that favors user-specific personalization of advertisements, and the other that focuses entirely on user input. Both cede some control to the web user, but his/her “choice” is a bit more direct in the latter model, especially as used by Clusty or Grokker.
We might also see this as a bet between who chooses the ads -- advertisers or consumers. And consumers do respond well to increased choice. Moreover, whenever consumer data is collected, it’s worth asking why and how -- which is another way of saying: maybe putting the “consumer in control” lowers the need to personalize ads. Collecting information is all well and good, but we should make sure that we need what we’re accumulating. There probably isn’t an easy answer to this, and I would be curious to see how Grokker's or Clusty’s clickthrough rates compare to MSN’s. But it might be worth investigating.
There’s a delicate balance to be struck here between making advertising relevant and giving consumers control. Both are major strengths of online advertising in general, and both have served us well in the past. As search marketing is improved, it may just be a matter of keeping both in mind.
Isaac Scarborough is manager of market intelligence at Chapell & Associates. Read full bio.