• by Steve Smith , Staff Writer @popeyesm, April 11, 2008

We are years away from addressable forms of TV delivery, but until then there still are ways that cable and network programmers can employ some version of behavioral targeting to their medium. For instance, AMC now can attach the viewership to specific movie titles in its libraries to a wide range of product preferences and shopping habits. In much the same way online BT can root out fascinating unforeseen affinities between certain content audiences and products, Charlene Weisler-Schwarzkopf, senior vice president of research at Rainbow Media (Cablevision's unit that includes AMC) can see that viewers of the comedy "Caddyshack" may be perfect for financial services advertisers.  She tells us about the research behind such unlikely pairings.

Behavioral Insider: What basic data did you collect in this research about AMC viewers?  

Charlene Weisler-Schwarzkopf: The identity metric looks at individual film titles and matches the film preference and viewership to an entire questionnaire of attitudes, personality descriptions, purchasing decisions and behaviors. We actually ask more specific detailed behaviors against four major ad target categories: domestic auto, import auto, financial and theatrical.

BI: How did you gather the data?

Weisler-Schwarzkopf: We hired Nielsen Custom to help us construct a questionnaire that is 35 minutes long and online. It was 50/50 male/female, with a national sample size of 5,006. The age skew replicated the U.S. census, so we had the proper distribution by age. You had to have been a cable or satellite subscriber and you had to watch at least five hours of TV a week. We looked at AMC and select competitive networks.

BI: Were you actually finding out about individual audiences for specific film titles or working from buckets of genres and attaching preferences to those?

Weisler-Schwarzkopf: This started last year when we recognized that there were very distinct psychographic and behavioral differences when you collected it by genres. But we decided that there was a lot of overlap and it wasn't the perfect way of approaching it.  So we actually match individual titles. We included 150 movie titles that will or have appeared on AMC and we queried about preferences, viewership, enjoyment of those titles. And then we took all of these behavioral and psychographics and self descriptors and purchasing behavior and technology adoption and we matched it to the individual title.

BI: Were there any striking, unexpected affinities between titles and audience behaviors?  

Weisler-Schwarzkopf: What did surprise me is that when we went in for a particular financial advertiser for example, of course one might expect "Wall Street" and "Trading Places." But there were other titles that were included like "Catwoman" and comedies. Basically what this meant to me was that when people want to kick back and watch their favorite films they are not just watching a particular genre or movie title. They are watching a range of different things that seem to collect on a very psychographic level. So if you were to ask a person who might be the perfect target for a credit card what their favorite films are it may be across a compendium of different genre types. Yet they all connect on a very deep level to that viewer.

BI: So it sounds as if you can come at this now from two interesting angles. You have detailed product preferences and behaviors indexed by individual film titles and you can map products against some unexpected film properties.

Weisler-Schwarzkopf: We can go further. We built into this the ability to data-fuse. So now we can run a variety of different attitudes and perceptions for our films with Simmons. And we also data fused into Nielsen NPower. So we can post these titles on a minute but minute basis to show we can over-deliver in the actual target for the advertiser.

The way it tends to work is that a network goes to the advertiser and the advertiser says, we want adults 25 to 54. The buyers will say adults 25 to 54. The planners will say well, we want somebody who goes to the movies on the opening weekend and is a frequent moviegoer and rents a lot of DVDs. So you have adult 25 to 54, but then you have all these behaviors that become attached to it.

Because we have asked these questions in our questionnaire by movie title, we can then go through our inventory of films. We will pull out those titles for that age and gender break and for those behavioral breaks, and we will come up with a carefully curated list of films that we will sell to the advertisers. And then we will be able to say, we will guarantee that you will get X number of impressions based on this collection of films we chose for you because of your behavioral target. And we can run this information in Nielsen and be providing you with a Nielsen delivery level based on what we promised.

BI: Let's work out of an example of how some film titles map against a product category.

Weisler-Schwarzkopf: One is for finance. What tees to the top of the list for buying target of men 25 to 54 financial advertiser? The planning target is, he pays for financial advice, wants secure handling for finance online, and has a high importance of insurance. The titles that resonated were "Wall Street," "Any Given Sunday," "Apollo 13," "Caddyshack," "M.A.S.H.," "Ocean's Twelve" and "Batman Begins." And then we would package this for the advertiser with special marketing requests like celebrity or hospitality and some AMC specials. So it is going to be a very rich type of package for this financial advertiser.

BI: Are you already seeing, in talking to advertisers, specific categories that perhaps have not advertised on AMC being brought in.

Weisler-Schwarzkopf: I have taken it out to many of the top researchers at the agencies and have gotten some very encouraging response. I know that fairly soon we will be rolling this out in a big way to the buyers. But already we are seeing from our internal research that we are able to attract a new group of advertisers.

BI: Does this also have the possibility of affecting on-air programming in the way you schedule and program clusters of titles?

Weisler-Schwarzkopf: It is a very interesting question. It could be a very compelling way of packaging our inventory so that it really resonates with a particular group at a point in our schedule.