Ad Meter is a great cultural artifact which has been around for 20 years. It uses a live panel to record second by second sentiment about Super Bowl commercials. It works like this: 234 panelists use a device to continuously rate how they feel about ads in real time. An ad is given its highest average score at any point (second) in the length of the commercial by these panelists. I wonder why they take this approach (an ad with high highs and low lows can still win). Local avails are excluded from the eligibility for obvious reasons. Whatever you might feel about the method, it’s a uniquely fun and accessible diversion that is a loose gauge of the mainstream meme.

As you might guess, the highest rated ads are those which are most entertaining, tell a compelling story and engage the viewer. Like a show within a show. The top five from 2008:

Super Bowl commercials are the zenith of branding advertising venues and stand at the opposite end of the spectrum from search engine marketing. Of course both ends have a role and purpose in a marketing portfolio. Different spend allocations make sense depending upon the nature of a product category. In the SEM and direct response world, popularity and ratings are easy to measure - they’re measured in ad performance. That is ultimately revealed by cost of acquisition (COA).

But as brand advertising acquires some of the characteristics of direct marketing approaches such as SEM (and vice-versa), I’m left to wonder, is this the best we can do? I wonder why no one has disrupted the idea of Ad Meter (within a broader context).

Among the possibilities, I see Ad Meter changing over time on (3) attributes:

–> Broader Sourcing — In the future, anyone who wants to participate in rating commercials will be able through definitive channels. That is sort of the idea behind Nielsen’s approach to measuring the reception of Super Bowl commercials via its online voting site. But this quick-and-dirty social experiment does not have the same authority of Ad Meter. In order to achieve the same quality of participation online as the high-touch model being employed by Ad Meter, some problems will need to be addressed with creative solutions. Program-level innovations include identity management and user interface design. As a reminder, we are talking about brand advertising here since these problems are less relevant for the direct response business. A simple word on each point. In terms of identity management, it’s essential to authenticate users. This will require registrations (incentives to do so) and software-verification processes. On the user interface side, it’s not as straightforward as thumbs-up or thumbs-down. Deeper feedback and involvement introduces the need for more creative user interactions that focus on usability across different dimensions of feedback.

–> Measures of Effectiveness - Ad Meter currently uses an approach that allows participants to dial-up or down how much they like an ad throughout its run. This is a smart approach that provides the simplicity necessary mine impulses. But what are the things that factor into like and dislike? And are there simple ways of bringing those into the mix to drive more intelligence into the ratings. Three factors that might make sense: Originality, Engagement. and Virality. For Originality, we mean is this something truly creative. By Engagement, we look to measure the degree to which the content and storyline is being followed. When a room falls silent and you can hear a pin drop, that’s usually a sign of engagement. By Virality, we want to know whether viewers will proactively and naturally spread the word about the content because they want to. These ideas build on and complement the concept of unaided awareness, which is a way to know more about brand marketing effectiveness.

–> Creative Participation - The words people use to provide feedback might be just as useful and more informative than the movement of dials on a ratings meter. Words are more expressive but harder to quantify. But blogs are big and with things like Twitter and Jaiku, it’s not hard to imagine an instantaneous aggregation of commentary (unlike this blog post, which is a day late, a dollar short). A second form of additional participation would be to discover the kinds of things that viewers would like to consume to continue the experience of a good commercial. A simple approach would be to allow feedback on the usefulness of different promotional tie-ins (posters, screen-savers, follow-on content) or to make that available in exchange for registration.

So who might be logical candidates to drive such disruption? There are traditional suspects including online ad giants (Google, Yahoo, AOL) looking to grow their service lines and media measurement firms like Nielsen. More surprising candidates include a new batch of startups in the space of sentiment analysis. These firms use text mining techniques to measure opinions in the blogosphere related to brands and products. I recently ran into a friend who leads development at a sentiment analysis firm called Biz 360. They seem to be doing very well. A similar company is Umbria.

A big opportunity for these firms might be to use sentiment analysis to optimize ad delivery and participate in the upside of better performing ad placement strategies. In other words, to take revenue shares for their added value.

Back to the big picture….what these (3) dimensions of change have in common is the removal of friction from measurement, participation and response behavior of viewers. Friction remains central to high-touch, central processing approaches like Ad Meter. It will be nice to see how those barriers are overcome in coming years.

• Ad Meter 2008 Homepage from USA Today
• Full List of Ads and Performance
• Nielsen's Take on Ad Ranks
• Adweek's Super Bowl XLII Site
• Game 17 vs. Commercials 3
• Apple's 1984 Classic

Filed under: Online Advertising, Search Marketing, Web Apps, Traditional Media

Email It   Digg This!   Slashdot It!

Tags: Ad Meter, Ad Bowl, Bowl 360, Brand Advertising, SEM, Sentiment Analysis, Brand Advertising Performance,