Tom Webster, writing and speaking

Clicks, Cakes, And The Limits Of Social Media "Science."

Added on by Tom Webster.

There is an apocryphal story in the annals of market research that I particularly love about cake mix ("apocryphal," by the way, is Greek for "a pile of crap," so this probably isn't true - but I'll tell the story anyway.) The story goes that back in the 50's, Betty Crocker had developed its first completely one-box cake mix - just add water and bake. After some initial buzz, sales began to disappoint, so the Betty Crocker executives did a series of focus groups to suss out the problem. Imagine tackling this problem today, using social media monitoring, or tracking clickstream behavior. Betty Crocker might observe fewer clicks to their recipe page, or perhaps fewer positive mentions. Coupon activity from register scans might decline. Positive sentiment for Duncan Hines might increase. We might learn that the best time to tweet about cakes is 10:00 am on a Sunday. Maybe we'd record an increase in the number of tweets about the poor quality of Betty Crocker's mix.

We could take all of this online behavior - all of these tweets and clicks - and determine a few things. Some of our conclusions would be correct, while others would be off the mark. Mining this information is crucial to the lifeblood of the organization - don't get me wrong. But bits and bytes will only ever tell you the "what." They rarely give you the straight story on the why.

One thing I've learned in about 20 years of doing qualitative research - people are not as expressive about products and services as we'd like them to be. Often, we cannot clearly articulate what makes us uncomfortable, or dissatisfied, with a given product, so we fall back on the easy answers. "It doesn't taste right." "It costs too much." "I don't have enough time." These are the first things I hear in any focus group, before Stockholm Syndrome really sets in. This is when the experienced qualitative researcher reaches into their bag of tricks, and helps the respondents along - and uncovers the real reasons behind these perceptions of quality, value and importance. These data, of course, are anecdotal until you can test these assumptions, and social media is providing us with more and more tools to do just that. But social media often gives us the easy answers - not the true answers.

Back to Betty Crocker. Unable to mine Twitter, our 1950's executives did a series of focus groups with housewives that had tried, and ultimately rejected their cake mix. Much to their surprise, they realized that these ladies thought the cakes tasted just fine, and were pretty good values. Instead, the insight they developed over time was that the cake mixes were a little too easy. In postwar America, as their husbands worked long days, these stay-at-home moms were a little embarrassed about the fact that all they had to do to have a delicious cake on the table for their men to enjoy after work was just add water and stir. In short, they felt guilty.

This is why you now have to add an egg, or perhaps a little oil, to a cake mix. Certainly these ingredients could be incorporated into the package - we do have a little history of food science in this country. But adding these one or two ingredients made it feel like baking again, and not just assembling. These women didn't just want cakes - they wanted to feel good.

The numbers only give you half the story - and I say this as someone who makes his living telling the stories of numbers. The operative word there, of course, is story. It's easy to be seduced by social media data, especially by those who loudly proclaim that they have the numbers on their side. Numbers aren't on anyone's side. I've had a lifelong battle with them, trust me. Adding insights to data is more than just putting flesh on the bones of an otherwise solid skeleton. Often, you don't know what you think you know merely by dredging tweets.

No, data without insights is just ignorance.