Pepsi's Taste Test And Coke's Choke

It was the 1980s, farther away in years than in feeling. Coca-Cola was becoming increasingly concerned about its competitive lead over Pepsi. A decade earlier Coke held a four-to-one lead over their rivals when measured by "exclusive drinkers." By the early' 80s Coke held a one point lead over Pepsi even though they were spending $100 million more on advertising.

At that moment, Pepsi began a TV saturation-bombing campaign featuring beverage drinkers facing off in what they called "the Pepsi Taste Challenge." Dedicated Coke drinkers were asked to take a sip from two glasses, one marked Q and the other marked M. Which did they prefer? Invariably they would say "M" and sure enough, M was revealed as Pepsi. Coke's immediate disputation was to do their own head-to-head tests, yet they found the same thing: 57% favored Pepsi! A 57-43 margin is a lot, especially between two international soft drink Goliaths where millions of dollars hung on preference trending.

With seemingly incontrovertible forensics, Coke's leadership held a flurry of meetings, demoralized that Coke's mystique was waning. Coke had double the vending machines, more shelf space, while spending far more in marketing. Thus they concluded, they had better take the Pepsi Challenge seriouslyn said Brian Dyson, Coke's COO of domestic operations. His decision then led to an innovation dubbed New Coke, a sweeter more Pepsi-like taste. Immediately Coke analytics found preference shares rising, rounding up hundreds of thousands of testers across North America. New Coke actually beat Pepsi by six to eight shares. Execs were giddy, while New Coke production was given the green light. In his press conference the CEO cavalierly announced "This is the surest move the company's ever made!" How could New Coke fail? But it did ... disastrously. Through massive rebellion, Coca-Cola was forced to bring back original Coke. Enthusiasm for New Coke quickly died, but even more surprisingly, the apparent ascension of Pepsi's brand share never materialized!

What happened here? How could Coca-Cola be so wrong? Researcher Carol Dollard,, who worked for Pepsi for years, discovered that in a sip-test or CLT (central location test) tasters don't drink an entire can. In fact, they take a few sips and make their choice. But what if they had taken a large bottle home, then after a few weeks offered their opinion? One of the natural biases in a sip test is "sweetness." Dollard later opined, "many times I've seen a CLT give you one set of results but in a prolonged home test, you get a much different set of data." In short, Pepsi is a drink built for a sip-test, sweeter than Coke so out of the blocks Pepsi had an advantage.

Here we get into rocks and shoals because other less tangible elements factor in. For example, the phenomenon of "sensation transference" as defined by mega-marketer Louis Cheskin. Put in simple context, Cheskin believed that when we buy a product at a store -- including beverages --at an unconscious level most of us don't make a distinction between the "packaging" and the "product." So, the brand is the packaging and the product itself. In his early book "Blink," the brilliant Malcolm Gladwell writes about uncanny cause-and-effect in our daily lives including soft drinks and those findings surely extend to radio sampling and impressions.

Think about it: In the blur of programming to fickle and distracted Nielsen respondents, can we really afford to be cavalier about our positioning, music accuracy, or our talent's appeal and retentive value? What if we routinely did Pepsi-like intercepts where we paid respondents to hear a 60-second montage of station Q and a montage of station M? Would those "ear-tests" align with Nielsen monthlies? If you think so, call us to claim your free case of New Coke. What we must accept is our need to influence listeners' attention within very short time spans ... so make it sweet.