Debunking T.V. Advertising Myths

Leonard Lodish, from the University of Pennsylvania, led a team that explored T.V. advertising and learned that level of spending for ads does not correspond to sales levels. We thought it did. They also learned that good results with recall and persuasion tests, telling you that people remember your ads and believe your products are good, don't lead to increased sales. We thought that was true, too.

Lodish's team examined data supplied by Information Resources, Inc. (IRI), a Chicago firm, which has an ongoing research program. They identify typical U.S. cities, select five communities that meet their requirements, and receive scanner data from the grocery stores in these cities. Next, they select households and give the adults cards to present when they shop. Checkers scan these cards along with their grocery purchases, and the data goes to IRI. Finally, IRI gains control of T.V. commercials seen in these households by working with local cable T.V. companies and uses this control to conduct advertising experiments.

For example, IRI may test the effect of a new soft drink commercial by splitting their households into 2 groups and showing it 10 times a week to one group and 50 times a week to the second group (a test of advertising weight). IRI monitors purchases of this soft drink made by these households for up to a year and records differences between the 2 groups. If the more frequently seen T.V. ad fails to increase sales for the second group, then the soft drink company would waste money advertising nationally at the heavier rate.

Lodish, et al., examined 389 recent experiments like this, and they quickly realized they'd made a dramatic discovery: 61% of current T.V. advertising fails to stimulate sales, and this arouses our worst fears about advertising, that our ads are part of the useless 61% and we're wasting our money.

Recognizing that the rules for T.V. advertising have changed, Lodish's team returned to their data and searched for patterns that would reveal rules that work today. Here's what they found:

1. T.V. ads work better for products less than 3 years old, and prime time is best. Tests showed 55% of these ads stimulated sales.
2. T.V. ads work better for small, less well-established brands lacking clear marketing images.
3. T.V. ads intended to change existing attitudes are more successful.
4. T.V. ads for frequently-purchased items, lower-priced items, items in product categories with a wide variety of choices, and items in growing product categories are more successful.
5. Changes in T.V. ads which keep the message fresh and new stimulate sales, but when a T.V. ad fails to stimulate sales, increasing its frequency or increasing the length of the ad campaign causes a decline in sales.
6. T.V. ads in product categories that already have a great deal of T.V. advertising don't increase sales.
7. In-store merchandising promotions quash the effect of T.V. advertising.
8. T.V. ad campaigns are more successful when they reach out to new users and strive to increase usage among established customers (for example, pointing out a new use for a product).
9. And finally, T.V. ad campaigns for new products are more successful when ads are run more frequently at the beginning of the campaign. This tactic stimulates trial usage.

Lordish's team suggests manufacturers create their own lead markets, and carry out advertising experiments similar to IRI's program. Here's how: 1) Select a relatively self-contained community in your market and note the current level of sales. 2) Eliminate all T.V. ads for this area. 3) Monitor sales levels for 6-12 months, and if they don't decline, eliminate T.V. ads for your whole market while continuing to monitor your lead market. It will alert you with falling sales if you need to resume T.V. ads. When sales do fall, use this 6-12 month cushion to test commercials, and when you find one that increases sales, you can use it with confidence in your entire market. It is the results of these experiments that retailers should ask to see when asked to devote more shelf space to a manufacturer's products.

Reference: Lodish, Leonard, Magid Abraham, Stuart Kalmenson, Jeanne Livelsberger, Beth Lubetkin, Bruce Richardson, and Mary Ellen Stevens (1995) How T.V. Advertising Works: A Meta-Analysis of 389 Real World Split Cable T.V. Advertising Experiments. Journal of Marketing Research, 32 (May), 125-139.

Back to home page