TV Advertising Is Usually Not Worth It
Companies spend vast sums on commercials, but it’s been difficult to gauge their effectiveness. A new study offers a more reliable method—and some bad news for many brands.
Michael Meier
Organizations pour a huge amount of money into running ads on traditional TV networks—about $66 billion in 2019, according to one estimate. They must believe that such costly marketing efforts are driving sales. Are they right?
According to a new study, the return on investment for TV ads is dismal.
Anna Tuchman, an associate professor of marketing at Kellogg, and her collaborators analyzed the effect of TV commercials on sales for more than 200 consumer packaged goods, including food, drinks, and basic household products. The researchers found that the ads had a much smaller impact on sales than previous studies had estimated. For many products, the return on investment (ROI) was negative: the companies had spent more on commercials than they earned back in additional sales.
“It looks like the vast majority of firms are overadvertising or spending too much on advertising,” Tuchman says.
Questionable Returns
Despite all the money that companies funnel toward TV commercials, those companies lack reliable methods to quantify how well the ads work.
Many managers “admit that they don’t feel super confident in their ability to measure the effectiveness of that advertising,” Tuchman says. “It’s this big investment with questionable returns.”
Over the years, researchers have attempted to quantify the effectiveness of most advertising. Some of their analyses were case studies, in which the authors focused on one industry or product. However, it’s hard to generalize the results beyond those specific examples, Tuchman says.
Other researchers performed meta-analyses: they gathered many case studies and drew conclusions about overall ad effectiveness. But this method might be biased, Tuchman says, because researchers might have been more likely to publish case studies with positive results showing a strong effect on sales. Since the conventional wisdom in the field is that ads do boost profits, a case study that finds the opposite might arouse skepticism and remain unpublished.
There is a study from 1995 where researchers conducted a series of randomized control trials to analyze the effect of TV advertising on sales for many products. But those data were collected in the 1980s, and the effectiveness of commercials may have changed since then. “We want to update the results,” Tuchman says.
’Tis the Season?
Tuchman and her collaborators, Bradley Shapiro and Günter Hitsch at the University of Chicago, focused on 288 consumer packaged goods. The products were well-established—for example, Diet Coke and Bounty paper towels. Even though these brands were already household names, the companies were still buying a lot of TV ads: firms in their sample spent a median of $10.5 million per year on commercials for each product.
The researchers obtained data from Nielsen on the products’ sales at about 12,000 stores in the United States, as well as data on purchases by more than 60,000 American households, from 2010–14. The team also examined Nielsen data on traditional TV ads during the same time period. (Streaming services were not included in the data set.) For each commercial, Tuchman and her colleagues calculated the percentage of households in a particular local market that saw the ad.
Then came the tricky part: teasing apart the effect of the commercials from other factors that might influence sales. While firms did vary the number of commercials they ran throughout the year and the number of commercials they ran in different parts of the country, the researchers couldn’t just assume that any corresponding increase in sales was driven by the ads.
Companies might advertise more during seasons when sales tend to be higher anyway; for instance, a sunscreen firm might air more commercials in the summer. And perhaps firms were more likely to run ads in areas where their product was already popular.
To address the first problem, the researchers accounted for the effect of seasonality in a mathematical model of advertising and sales, which allowed them to determine excess sales that went beyond the typical seasonal bump.
The team also examined customers who lived near the borders of TV ad markets. Commercials are aired within geographic zones called designated market areas (DMAs); the country is split into about 200 DMAs. People who live on opposite sides of a DMA border tend to be similar in terms of product preferences and other factors that may affect demand. So if a firm ran a commercial in one DMA but not the neighboring DMA, the researchers could check whether sales in the ad-exposed side of the border rose.
Small Effects
To determine how much commercials drove sales, Tuchman’s team calculated a measure called advertising elasticity, which captured how much sales changed with a given increase in ad exposure.
The median elasticity across all products was 0.01—which meant that if a firm doubled its TV ads, sales would rise by only 1 percent. This figure was an order of magnitude lower than several estimates from previously published meta-analyses; those studies had suggested sales increases ranging from about 9 to 24 percent for every doubling in ads. One possible reason for the disparity is the bias toward publishing case studies with positive results, Tuchman says.
If the TV is on, the viewer is counted as being exposed to the ad, “but you didn’t actually even notice it because you were so absorbed in your phone.”
— Anna Tuchman
Tuchman and her colleagues’ result was also lower than the elasticity found in the 1995 randomized control trial study, which had estimated a boost of 5 percent for every doubling in ads among established products. Tuchman speculates that this difference may be due to the fact that consumers today might not pay as much attention to TV ads. Many people now read their phone or tablet during commercial breaks. If the TV is on, the viewer is counted as being exposed to the ad, “but you didn’t actually even notice it because you were so absorbed in your phone,” Tuchman says.
The researchers also estimated how many brands saw any positive effect of TV ads on sales at all. For roughly two-thirds of firms, “these effects are essentially zero,” she says.
The team then wanted to calculate ROI—that is, how much firms actually profited from the advertising after accounting for the expense of buying the ads. This analysis presented some challenges because it required two additional pieces of data: the cost of each ad and the profit margin on each item.
The team estimated ad expenses based on average-advertising-costs data. They also calculated ROI for each brand multiple times, considering a range of realistic profit margins based on industry reports.
Regardless of which profit-margin value they used, the picture looked bleak. The weekly ROI was negative for more than 80 percent of products.
“It looks like firms are really overadvertising, and they would be better off reducing their advertising spending,” Tuchman says.
Smarter Marketing
Given these dismal returns, how could firms not be aware that the money they lavish on commercials is largely wasted?
One possible explanation is that employees in charge of managing TV commercials have no incentive to show that their ads don’t work, Tuchman says. If you do it, “you put yourself out of a job.”
Companies might also lack sophisticated methods to measure the effect of ads. For example, a sunscreen company might attribute more of the increased summer sales to commercials than it should.
To overcome these hurdles, firms could set up an independent data-science team to analyze ad effectiveness, Tuchman suggests. And companies should consider cutting TV commercial budgets in favor of other marketing strategies.
The study focused on established products, and the results may not apply to newer brands, Tuchman notes. People might pay more attention to commercials for a product they’ve never heard of before. “My guess is that new products would see larger ad effects,” she says.
