TV: As Effective as Ever

The drumbeat of doom for TV advertising has sounded for more than a decade—DVRs, channel surfing, fragmentation, clutter, the flight to digital media … Jay Leno moving to primetime. Now the recession has even TV’s most reliable moneybags of yore, such as Procter & Gamble and General Motors, yanking big wads of cash off the table.

Yet a funny thing is emerging from the smoldering ruins of what may be the ugliest quarter TV has ever encountered financially:

A growing body of evidence which suggests not only that TV advertising still works, but that it may be working better than ever.

Analyses by people and companies that have studied or made bets on advertising effectiveness for years find no evidence that all of the problems TV advertising faces have done anything to render it less effective at its ultimate goal—selling stuff.

Old TV

A seven-figure ethnographic study due to be released next month by the Nielsen Co.-funded Council for Research Excellence from research firm Sequent and the Center for Media Design at Ball State University appears set to punctuate that point, finding that TV remains the dominant medium even for reaching youth, despite the inroads of digital and social media, according to a person familiar with the research. That just adds to findings by people who have been researching or using TV all along.

If time shifting, ad skipping or clutter really were rendering TV less effective, then it should show up in marketing-mix analyses that have been done since the early 1990s as a lower average sales lift per gross rating point over time.

It doesn’t, according to Marketing Management Analytics (MMA), a unit of Aegis Group’s Synovate, which has done such analyses for longer than anyone. “To be honest, we haven’t seen a significant trend in the erosion of effectiveness of TV,” said Douglas Brooks, senior VP of MMA. In fact, MMA, which reports to clients each year on its findings regarding aggregate TV effectiveness, has seen a slight uptick in effectiveness in recent years.

MMA also has found a surprising spillover effect for TV in digital media—specifically that about a third of search queries for brands studied are driven by offline advertising, particularly TV—a higher proportion than that driven by online display advertising, Mr. Brooks said.

Leonard Lodish, a marketing professor at Wharton and one of the authors of the 1995 “Why Advertising Works” study has discovered equally surprising results as he’s worked to update that research. He’s found that TV advertising actually became more effective, not less, after 1995, in a paper published in 2007 by the Journal of Advertising Research, soon to be updated in a new study now awaiting publication by the same journal.

He got at the findings differently than MMA, and with less statistical modeling required, by using data from Information Resources Inc.’s BehaviorScan markets and other matched-market tests that compared different levels of spending in different test markets. Specifically, the average volume lift from incremental TV spending has increased since 1995, according to the study by Mr. Lodish, Wharton colleague Abba Krieger and University of Houston marketing professor Ye Hu.

One reason could be that commercial avoidance, fragmentation and clutter actually increased the reward from spending more—or at least above a certain noise threshold. But the study also found a similar, if smaller, improvement since 1995 in volume lift for brands when they had any amount of TV vs. having none at all.

Despite the improved effectiveness since 1995, more than half the advertisers (16 of 29) in the study still lost money by running their TV ads. “The ones that did make a profit, though, did very well,” said Mr. Lodish, to the extent that on the whole, advertisers in the study made a profit from their TV ads after 1995, but lost money before 1995.

That’s obviously a serious caveat to the value of spending on TV. The other caveat is one that other marketing-mix analysts also report from client work that creative quality makes a big and increasing difference, in many cases explaining more about success and failure than media choices.

Mr. Lodish said he still doesn’t really know how TV advertising effectiveness could have increased since 1995. Mr. Brooks can’t really explain it either, though he has a theory that, while self serving, may be true: The highly analytical clients using marketing-mix modeling or matched-market tests may compensate for the impact of DVRs, fragmentation and clutter by making smarter bets.

“When the fish get finicky,” he said, “it makes you a better fisherman. The presentation of the bait and how it’s delivered—getting it in the right spot at the right time—becomes critical.”

The other question Mr. Brooks often hears these days is whether recession historically has caused the average sales lift per GRP to decline. The answer, at least based on data from the relatively mild recession of 2000-2001, is no.

Maybe that’s because share of voice goes up and clutter goes down for advertisers who stay the course, he said. Or it may be that advertisers who stick around focus even harder on only running ads that work.

Looking for even more evidence? Executives at A. Eicoff & Co., a unit of WPP Group’s Ogilvy & Mather that’s one of the biggest and oldest direct-response TV shops in the U.S., said response rates from its TV ads haven’t deteriorated at all over the years.