Pepsi’s Digital Screw-Up

Posted by on August 11, 2014

Todays memo is a long one, but I promise you it’s worth it.

Advertisers are attracted to online media when they’re not entirely happy with their investments in traditional broadcast media. To understand the reasons behind their disappointments, we need only to revisit the subject of last week’s Monday Morning Memo:

“Linear, no-threshold thinking” assumes that every statistic is scalable. It’s what causes advertisers to assume they can “test the waters” with small investments, then increase their financial commitments if the test results are positive.

If an ad needs to be encountered only once to trigger a sale, it’s a direct-response ad. Congratulations! You’ve successfully crafted a high-impact offer for a product with a short purchase cycle. Direct-response ads are scalable, meaning sales increase proportionately to the number of people reached. But not everything can be sold with a direct response ad. The simple truth is that most products and services require that their ads be encountered again and again.

Pepsi has been a household word since before we were born, so why do they keep advertising? Couldn’t they reduce their mass media spending and still maintain their sales volume?

In a word, no.

We know this because Pepsi tried it.

Bob Hoffman was the keynote speaker at the 2014 European conference of AdvertisingWeek:

“In 2010, Pepsi cancelled all its TV advertising and its Superbowl advertising to great fanfare and bet BIG on the largest experiment in social media marketing ever attempted, ‘The Pepsi Refresh Project.” TIME magazine quoted the CEO of a New York brand consultancy, ‘This is exactly where Pepsi needs to be. These days brands need to become a movement.’ Well, they became a movement all right. I estimate The Refresh Project cost them between 50 and 100 million dollars. It got them 3.5 million Facebook likes and a 5% loss in market share, which they seem to have never recovered. That year, they dropped from the second best-selling soft drink in the US to third. Pepsi’s marketing director said, ‘The success has been overwhelming. We have more than doubled our Facebook fans. We have more than 24,000 Twitter fans.’ The L.A.Times didn’t see to agree. They called it ‘a stunning fall from grace.'”

Hoffman went on to say that TV and Radio are best at creating demand, while the web is terrific at fulfilling demand. The interviewer then challenged Hoffman by saying, “But it is changing. And it’s changing fast. Ten years ago 93 percent of the public got their news from television and only 7 percent got their news online. Today it’s 26 percent online.”

Hoffman’s response reflected his 40 years of experience directing ad campaigns for McDonald’s, Toyota, Shell, Nestle, Blue Cross, Chevrolet and Bank of America:

“What we often confuse is the use of digital media with its power as a marketing or advertising entity. The fact that more people are using online for news is not a de facto proof that it’s a good advertising medium. Let me give you an example of that: the old-fashioned telephone. Everyone in the world had a telephone. It was a hugely popular means of communication. That didn’t make it a good advertising medium. It was a lousy advertising medium. The fact that people us it for communication or to get information or to have conversations doesn’t necessarily make something a good advertising medium.”

Now let’s get back to the subject of why so many advertisers are frustrated with their TV and Radio campaigns.

In last week’s memo we described motorcycles going out of control when their riders accelerated them beyond the “safe speed” threshold while navigating an S-curve. Trips through the curve below the safe threshold speed are uneventful, but trips through the S-curve above the threshold are dangerous. In other words, the ratio of crashes to speed isn’t “scalable” because the motorcycle behaves very differently at speeds above and below the threshold.

The skill of the rider is another variable, of course, but although skilled riders might navigate the curve at higher speeds, there’s always a threshold at which even they are going to crash.

Thresholds are inevitable when measuring human response.

We must also keep in mind that humans attach complex meanings to sound. This is what makes TV and Radio effective at influencing people who aren’t currently, immediately in the market for your product or service. TV and Radio win the heart’s preference, then patiently wait for the customer to be ready to buy.

The motorcycle safety threshold is all about

(1.) speed and
(2.) the skill of the rider.

But mass media advertising is all about

(1.) repetition and
(2.) the impact of the message.

TV and radio campaigns that deliver minimal results in the first few months often become highly effective when they’ve crossed the repetition threshold of the listener. A customer needs to encounter the average message multiple times before it is likely to be retained.

Advertisers often ask, “How many times does the average person have to see or hear my message before it will be transferred into the automatic recall part of the mind?” Although this seems like a reasonable question, it’s a little bit like asking, “How many ounces of alcoholic beverage does it take for the average person to get drunk?” We can’t really answer that question until we know whether the “ounces of alcoholic beverage” are beer with 5% alcohol, wine with 14% alcohol, or Scotch with 45% alcohol.

How strong are your ads?
The stronger your ads, the fewer times they have to be heard. And even then, as Pepsi learned, the customer will sober up and forget you if you leave them thirsty long enough.

Strong ads are created by strong writers.
How many do you have working for you?

Roy H. Williams

PS – We’re going to look much, much deeper into all this during today’s session of Wizard of Ads LIVE (Aug. 11, 2014) Details are available at

Leave a Reply