• by Frank S. Foster , August 14, 2008

Five years ago, I had the opportunity to work with a number of talented researchers in the analysis of data obtained from one of the early addressable television advertising field trials.  One thing that became clear early on was that addressability affected channel surfing.  By examining how a set-top box typically behaved during non-targeted commercial breaks, we could establish a commercial-tuning baseline.  We then compared tuning behavior during addressable commercials and saw significant deviations from the baseline. 

Whether the baseline was averaged over a period of time, by genre, by daypart, or by network, the end result was the same.  Addressable technology often prompted the person controlling the remote to significantly change his or her ad-skipping habits.  While not unexpected -- if the message is relevant to a viewer, it follows that they are more likely to stay tuned -- the quantitative research was reassuring.  While penetration of time-shifting technology was low at the time, technology allowing viewers to skip ads was making many advertisers and their agencies nervous.  Even though the trial involved only linear viewing, addressability had the potential to be a weapon advertisers could use to fight TiVo's fast-forward button. 

Five years ago, I thought the industry faced many hurdles before the technology would be ready for mainstream deployment.  Some of those issues have been overcome but several remain, including:

1)     Bandwidth optimization

2)     Content creation

3)     Audience targeting

 

In the trial where our data was collected, four different broadcasts were assembled for each insertable network.  The network programming for each broadcast was identical, but the commercials were grouped based on demographic targets.  Thus, a set-top box would be mapped to a specific version of the network tailored to a rather generic demographic profile.  While acceptable for testing purposes, the required bandwidth is completely unacceptable in today's environment.  I expected that new multiplexing technologies and more innovative approaches to the addressable problem would fix the bandwidth issue, but I have yet to see an elegant, scalable solution.

Five years ago in the trial, precious little consideration was paid to creative by many of the participating advertisers.  It appeared that individual commercials were selected for specific demographic targets from a pool of available inventory.  In my opinion, none of the commercials were created with addressability in mind.  With enough raw footage and a little planning, it's  not that much more expensive to assemble multiple versions of the same commercial.  I thought then that a few of the large advertisers might embrace the concept, but that too appears to be a hurdle in front of the industry, not behind it. 

Even though content creation remains a knotty problem, I said the same thing five years ago with respect to Web site development and blogging.  Take a look at www.squarespace.com.  In one evening, anyone can assemble a professional looking Web site that is easy to update, expand and most importantly, requires no coding expertise.  Data collection, blogging, html, video -- the service has it all.  While it is a much more difficult task to make high quality video production tools, it only takes one enterprising company.  But the question remains, will it happen sooner rather than later?

Audience targeting is a more difficult dilemma.  Acxiom and Experian seem to be getting experience by participating in a number of trials around the country.  While demographic and product consumption data are certainly interesting to a number of advertisers, behavioral targeting -- a technique used by online publishers and advertisers to increase the effectiveness of their campaigns -- has largely been ignored by those in the television industry. In my opinion, it is this connection with behavior that will open the door to success for addressable advertising.

Others including Steve Farella and Audrey Siegel, who wrote a piece in Media magazine this month entitled, "Why the Mix Doesn't Matter," agree that understanding viewer behavior is essential.  Farell and Siegel  suggest that the average ratings and reach potential of cable networks have become increasingly alike over the last five years and that the "proper" network mix does not insure success.  Consumer connection is their mantra.

Enter behavioral targeting.  Similar to the online environment, such an approach for television would use tuning information collected from set-top boxes in lieu of web browsers.  Data such as networks, programs or commercials tuned, and the interactions with the electronic guide, would be used to select specific commercials.  Want to reach a potential customer during the late-night news -- but only those with children between the ages of four and 10?  Target those whose history includes five hours a week of Nickelodeon viewing.  Advertisers could further refine their approach by combining behavior requirements with geographic or demographic constraints.  The possibilities are intriguing.

So why has behavioral targeting not caught on with cable operators and advertisers alike?  Privacy seems to be a big concern.  But the Cable Communications Policy Act of 1984 has a carve-out specifically for marketing and improving the customer experience.  Besides, behavioral targeting can surely be conducted in whatever privacy-compliant fashion the law or public perception demands.  Experian stores credit report details on nearly every person in the country.  If they can be trusted to give that information out only to those who need it, why can't they be trusted to give advertisers access to our television viewing data? 

In the end, the advance of addressable advertising seems to be moving at a glacier-like pace across the television landscape.  Let's hope the operators are not waiting for a transition to an IP infrastructure before they make the leap.  Otherwise, we will be seeing another column like this one five years from now.