@mattyglesias @Noahpinion I worked for the market research firm that invented the all-time greatest real world laboratory for testing TV advertising's effect on supermarket sales. Famous firms loved us ... for about 5 years. But only rarely did we find a statistically significant effect from more ads.
1/x
1/x
@mattyglesias @Noahpinion The one test I ran where more advertising moved the needle of sales as measured by supermarket scanners was when a client made an important breakthru by adding a highly effective new ingredient to their toothpaste. More (informative) advertising sold more of the new toothpaste.
@mattyglesias @Noahpinion But most of the time, brand managers hired us to prove to their bosses that their ads were so great that their TV ad budget should be doubled.
But that almost never worked (unless they had new news about the product to convey).
2/x
But that almost never worked (unless they had new news about the product to convey).
2/x
@mattyglesias @Noahpinion 3/x. It wasn't just that our real world lab didn't show that doubling TV ad budgets for well-known supermarket products didn't pay off. Almost all the time, it didn't have _any_ impact on sales at all. The 3000 families in the More Ads group would buy 0.1% less than the Control.
@mattyglesias @Noahpinion 4/x
After 5 years of spectacular growth in demand for our superb real-world TV ad lab, the big consumer packaged goods firms got bored with testing more TV ads with us and our sales declined steadily, and the business was finally shut down after a couple of decades.
After 5 years of spectacular growth in demand for our superb real-world TV ad lab, the big consumer packaged goods firms got bored with testing more TV ads with us and our sales declined steadily, and the business was finally shut down after a couple of decades.
@mattyglesias @Noahpinion We ran one test where 3000 families saw on their living room TVs twice as many ads for a famous brand, 3000 saw the normal budget of ads, and 3000 saw fewer ads. All 3 groups bought the exact same amount of the product.
@mattyglesias @Noahpinion I suggested to the big client that they test with us a strategy of cutting their advertising budget to save money. Start testing in our service 2 years before you roll lower budgets out nationally, so we can tell you if cuts finally start to hurt.
@mattyglesias @Noahpinion The client replied that as a corporation they believed in television advertising. They had gotten to be the world's most respected consumer packaged goods firm by massive advertising so they weren't going to consider cutting advertising no matter what our tests showed.
@mattyglesias @Noahpinion The incentives within the client encouraged ambitious brand managers to constantly demand larger TV ad budgets. The CEO had long ago launched his rise by getting the American Dental Association to endorse his toothpaste for having fluoride and advertising the hell out of that.
@mattyglesias @Noahpinion I worked on a couple of internal meta-analyses of the results of our tests: only rarely did more advertising move the needle of sales.
Anyway, it's all rather curious 1980s ancient history. I've often wondered about the larger implications, but I don't know what they are.
Anyway, it's all rather curious 1980s ancient history. I've often wondered about the larger implications, but I don't know what they are.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
