At the ANA conference a few weeks ago, we saw some great presentations from Target, General Mills, Chobani, Nissan, T-Mobile, Walmart and many other iconic brands and companies.
A few presentations really stood out for me, however it was the Kraft presentation by CMO Deanie Elster that was the most inspiring. Deanie demystified how Kraft built their “Big Data” machine and shared how Kraft’s big data is really comprised of small data. Essentially, “big data” is a collection from traditional and digital sources that represent a source for analysis, discovery and insights.
Deanie told us to start with little data. For example, Kraft boasts over three million recipe interactions per day on recipes.com and more than one million paid subscribers to its magazine Food & Family, which sits on a massive publishing platform. This platform is also producing boatloads of data, and two years ago the company made an official commitment to use the data to better understand its customers at the individual level.
For instance, in re-introducing its Philadelphia Cream Cheese business recently, Kraft was challenged with users walking away from the Philly brand. At that point, Kraft’s emergent platform “gave them the confidence to understand the different consumer segments,” Elsner explained. “Tying it to purchase data, we were able to define and locate where those consumers were. Understanding what motivated them between sweet and savory, we served up messages that were appropriate to the target in the right medium at the right moment.”
For instance, aiming at a consumer cohort Kraft labeled as “lapsed loyals,” Philadelphia’s message conveyed with relevantly targeted banner ads was “More Real Veggies. No Artificial Flavors.” And ads aimed at bagel buyers touted “More Real Strawberries. No Artificial Flavors.” The results of such niche targeting, Elsner said, “have been tremendous.”
Kraft also has taken a major plunge into more branded content, which is increasingly seen by American consumers as on par with content that they seek out from traditional publishers. Kraftrecipes.com, for instance, is a storehouse of 27,000 culinary professional recipes, 30,000 consumer-submitted recipes, and gets one billion recipe views a year. That’s good enough to rank Kraftrecipes.com as No. 7 among all recipe web sites.
Another example of big (and little) data came from Stephen Quinn, CMO at Walmart. Stephen shared that weather and sales of certain products are directly tied together. Like most other retailers, an umbrella or snow blower display in advance of rain or snow are very obvious in terms of merchandising. However, by tying in with weather.com, Walmart has developed a sales prediction model. Consumers like to eat steak when it is warm, windy and no rain is in the forecast. Salad consumption increases in temperatures above 80 degrees, but with low wind (would the croutons blow off the salad??). Ground beef sales spike with higher temps, low wind and sunny skies. Berries are consumed with temps lower than 80 and with low wind conditions.
Walmart has also tapped into social data. For example, when mason jars reach a critical mass on pinterest, Walmart capitalizes on that. They will devote an endcap to mason jars, and expect a significant lift in sales.
So, instead of being overwhelmed with big data, let’s start with small data. Take a look to see where the data relationships are with your customers, and then take action with the insights we develop from identifying these relationships.