Mining Data for Marketing Gold


When starting a project for a client, we routinely ask for back information on the brand to develop a foundation of existing knowledge and to avoid spending time “reinventing the wheel.” Back in the days when paper ruled, a brand that had been on the market for several years would amass shelves and shelves and boxes and boxes and stacks and stacks of information. Market data, target studies, sales and business data, advertising ideas, qualitative and quantitative research results, competitive intelligence and on and on. The client would invariably ask, “what specifically do you want to see.” We would say, “We want it all.” The client would counter with, “Are you sure? It would fill a truck,” and we would just ask, “when can the truck be here?” Today, for the most part the paper has been replaced with computer files and, if anything, the volume of back data has increased. But, the value of that information for our inductive analysis process has not diminished one bit.


Because, to us, that information is less a historical record of OLD information and more a treasure trove of NEW ideas and possibilities. We use the term data mining to describe the approach. We borrowed the term (perhaps ironically) from the computer world where data mining is defined as “the computational process of discovering patterns in large data sets involving methods at the intersection of artificial intelligence, machine learning, statistics, and database systems.” (Whew!) But our inductive, qualitative approach to data mining is a far cry from the statistically driven, numbers crunching process the name implies. Instead, it is a process of selectively looking for nuggets of knowledge and insight that can unearth new ideas for a brand or novel answers to pressing questions.

When is data mining something to consider? When you are looking to reinvigorate existing brands, find new opportunities in their target categories or just inspire a new line of thinking in your teams. For most strategic projects it is the first critical step. We have even had 3-phase insight/strategy or positioning projects where we have found a critical piece of the project deliverable in that first data mining step. There are also Clients who have requested stand-alone data mining projects with highly successful results.

Here are some of the things we’ve learned about getting the most out of existing data.

Set out with a purpose The whole concept of data mining is that the same information can drive different insights and conclusions when looked at through different lenses. We’ve often gone back to the same foundational documents and studies to apply the learning to a variety of situations. We’ve found that having the question or issue we are addressing “on the wall” as we go through the information helps us perceive and interpret what we are looking at in the most relevant way.

…or take a random walk Having said that, there are also times when we have gone into the data with absolutely no preconceived notion or objective in mind…on purpose These are most often the projects where we have been asked to look for new opportunities for the brand or for Clients looking to spark a new approach to the current marketplace. This kind of zero-based, free-form investigation is perhaps the most challenging…and rewarding. The “secret” is staying open to seeing patterns evolve, themes emerge and opportunities spontaneously “reveal” themselves. How do you know when you’ve found something of value? Try it. You’ll know it when you see it.

Ignore the cover page We have often been asked why we are poring over copy test results when we are looking for target insight or why we are reviewing target research in looking for a business-to-business solution. The inductive kind of process we employ in data mining is dependent on the volume and quality of the raw material we find…NOT the specific reason it was originally generated. That’s why we ask for the truckload. Taking the information OUT of context reveals its basic nature and worth as a building block for new insights and unique ideas.

Look for the “why” behind the “what” Often we’ve found that looking past the fact, rather than directly at it, is a technique for gaining new knowledge from the old. Some facts are very specific or numerical. Not easily open to re-interpretation. But, raising a specific fact to a general principle or thinking about what drove the number can reveal the new learning we are looking for. For example: X% of respondents found the message annoying, but the rest did not. What personality trait or behavioral characteristic might lead to that negative reaction? Was it specific to the articulation or was it tied to the product in general? What other evidence is there to help support or refine a personality profile as it relates to the product?

Make it a group effort Old joke – “How do you eat an elephant?” “One bite at a time.” That’s how we attack the “truckload” or file upload of data when it arrives. We divvy up the individual documents and each of us reviews one batch. Then we trade for another batch and do it again. And again. And again. Until we’ve each had a chance to look at it all. There’s a good reason for this beyond just making sure everyone is up to speed. And why we don’t stop when each document has been seen by just one of us. Everyone looks at information in different ways. Some of us are statistically driven. Some spot themes or general trends emerging. Others pick up on nuances of language and attitude. All from the same material. Having multiple eyes on every line of information helps guarantee the most robust pool of raw material for further development.

So, if the idea of data mining has piqued your interest, for today’s Strategy Break here are a few questions related to getting the most out of the data you currently own:
  • Is it possible to address this question with what we already know instead of setting out to do new research?
  • Can my team identify a new opportunity for the brand that is outside the scope of today’s brand plan without making a big investment?
  • What would be a meaningful project for the intern I’ve had assigned to my team that would be of real benefit to the brand, too?

After mining the existing data you might also want to take a look at our January, 2016 Strategy Break on Connecting the Dots. It provides some thoughts on how to work the data once you’ve extracted it from the existing information. Or give us a call. We’re always eager to roll up our sleeves and start digging.

Leave a Reply

  • (will not be published)

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>