I was surprised at the findings revealed by Dr. Christine Moorman, director of The CMO Survey.

The results of the most recent survey showed a decline in the number of projects in which companies use marketing analytics that are available and/or requested. According to CMOs surveyed, they report a 30% usage rate. This number is down from 37% a year ago.

“Big data” has been the rage for the past few years. CMOs recently reported that the percent of their marketing budgets devoted to big data will increase from 6% to 10%.

So while more is being spent on, and written about, “big data,” less of it is being used.

Recently in a LinkedIn discussion group, “Chief Marketing Officer Network,” someone posed the question, “Do you think ‘big data’ actually baffles most marketers?”

I think Dr. Brian Monger hit the nail on the head when he said, “I think ‘big data’ baffles a lot of folks. It’s why you not only need a good analyst to do the numbers, but an even better insight person to work with the analyst and the numbers.”

I know there are several graduate programs in business analytics; however, I don’t know that these programs teach people how to look for “insights” in analytics.

I had the pleasure of working on Procter & Gamble, Richardson-Vicks and Warner Lambert business early in my career. We received Nielsen data every two months on sales, market share and distribution by channel (food, drug and mass merchandiser).

There were also quarterly spending reports from MRI and other sources.

While some might not call this “big data,” my understanding is that it came from SAS and there was certainly enough to run single and multi-variable regression analyses that enabled me to determine:

  • The price differential at which Bounce Fabric Softener would lose share to generics.
  • The negative correlation between Wondra Hand & Body Lotion’s brand development index (BDI) and average relative humidity.
  • The positive correlation between advertising spending and Rolaids’ sales.

Subsequently I worked on Wachovia Bank for five years which fielded quarterly research to determine top-of-mind awareness, market share and switching preference. The availability of this data enabled me to show the positive correlation between switching preference and emotional versus rational advertising.

Most recently I was the director of marketing for the U.S. subsidiary of an Ireland-based manufacturer of wastewater treatment systems. When I joined the firm, there was a great deal of concern over the decline in sales since 2009.

By looking at sales and housing permits by month for the past 12 years, I was able to show a 95.6% positive correlation between housing permits and sales of their peat-fiber biofilter.

While this didn’t increase sales, it certainly explained the decline. This year, coincidently, sales are back up, along with new home permits.

With all of the data collected and available today, the ability to show positive and negative correlations, and to model future outcomes, is infinite.

However, you need to know what to look for. What insights do you want to get?

I recommend creating hypotheses about what is, or is not, driving customer behavior, sales, market share, or what ever else you want to know about your business.

Once you have created the hypotheses, you can then perform the data analysis to prove or disprove your hypotheses. The results of this analysis are insights.

What questions can data answer for you?




Tom Smith

Experienced marketing professional who has worked with more than 120 clients in 18 different vertical industries. ♦ Differentiate products and services by improving UX and delivering memorable CX to create an emotional connection to the brand. ♦ Obtain insights from analytics to solve business problems and drive revenue. ♦ Develop and implement marketing campaigns that double traffic and leads in three months. ♦ Certified Marketing Automation Professional ♦ Certified Voice of the Customer (VOC) Professional