• by Joel Rubinson , Op-Ed Contributor, June 24, 2008

A bad pun on a Sigourney Weaver movie, but let's face it, TV is still the 800-pound gorilla in the media mix. However, with the explosive growth of Internet advertising, viral/social community/event approaches, and shopper marketing, the questions concerning TV advertising's ROI have gotten more acute than ever. For example, at a recent Association of National Advertisers conference, a speaker suggested that 10 quality relationships might be worth more that 5 million impressions. At this point, I think the industry needs to engage in some serious math rather than qualitative characterizations to answer questions like that.

Some justify TV advertising as part of an integrated marketing approach, but given the size of the gorilla, it's hard for me to feel that TV's share can be justified if it is just a multiplier on top of the effectiveness of other communications efforts. TV advertising needs to be able to establish its payback on a "main effects" basis, and then synergistic effects become sweeteners.

While other media such as the Internet and shopper media are growing much faster than TV, it is, in fact, the largest media channel and is still growing so advertisers must have some confidence in its ability to deliver. However, much of this belief is probably faith (or perhaps fear) rather than fact-based.

Let's change that. I would like us to get at the facts via a meta analysis of what the industry has learned and hopefully trend over time regarding TV effectiveness as a commercial medium.

Let me get the ball rolling, but please, regard this as the start of the dialogue, not the answer; that will come from you, the marketing and media community with me, the Chief Research Officer of the ARF, simply reporting what we actually collectively already know but have never really harvested before.

To get started, let me present some of what I have learned.

* TV advertising response elasticities (percent change in sales as a result of percent change in advertising expenditures) were documented to be higher in more recent IRI BehaviorScan tests (conducted after 1995 vs. 1989-1995), as reported in a recent paper co-authored for the Journal of Advertising Research by Prof. Len Lodish. Furthermore, TV advertising exhibits long-term effects; the three-year cumulative sales effect of weight tests was calculated to add up to twice the impact that shows up in short-term analysis.
* At a recent ARF Council meeting, Marketing Evolution, Inc. reported on the results of 40 studies that revealed, on average, approximately 70% of the lift in consumer response was due to TV advertising which was in proportion to the budget allocation for the brands studied.
* Recently, Brightlineitv, an agency that focuses on interactive TV, reported a 5% "click-through" rate on TV advertising (% impressions that led a viewer to go to the corresponding interactive channel. That's huge as a market research finding regarding TV effectiveness, especially when we consider that many of the 95% must have gotten the message in some way even though they did not click-through.
* At a 2007 ARF media effectiveness council meeting, Art Christiani presented an analysis of Apollo data that suggested the industry might have been underestimating the impact of TV advertising by not having single-source data that would enable respondent-level modeling.
* On the other hand, ratings for leading shows are dropping dramatically (a third of what they were 20 years ago), commercial clutter has increased, viewers can fast forward shows on their DVRs, and CPMs are rising. It is no longer clear that a marketer can even achieve desired reach via TV alone.

The question I am raising is not if TV is the better alternative to other platforms...obviously, today, all marketing plans need to have multi-platform engagement and shopper marketing components.

While the prospects for TV might be on the rise given the advent of interactive TV, integrated internet, and behavioral targeting, this call to action is about current knowledge. It is about establishing the baseline of how well TV advertising works in the current environment rather than what might be in the future. I plan to take the lead at having the ARF reach out to organizations that have great expertise and continuous experience at modeling TV advertising effectiveness. I would like to conduct a meta-analysis of what is known about trends in TV advertising effectiveness over time. Let's get the facts!

What are your thoughts? Which of you will volunteer to share information with the ARF so we can advance the knowledge base on this critical subject?

Joel Rubinson is chief research officer of the Advertising Research Foundation.