Editor’s note: Lisa Brink is senior director, customer strategy at research firm Elicit Insights, Minneapolis.

Companies have such easy access to data today that they can either be paralyzed or liberated by it. Some believe big data answers all questions while others feel research provides a deeper understanding of behavior. Whatever your perspective, the truth is that there is more power in using them together to solve business challenges than leaning on one. These two important functions complement each other and can bolster insights. The big data vs. research mentality is dead, replaced by an understanding that these two sources of data used together can be transformative.

However, organizations aren’t always able to take full advantage of these functions. Analytics and primary research departments tend to have separate learning plans and objectives and they can live in different parts of the organization. Setting aside organizational structure challenges, these functions don’t need dotted lines to experience the benefits of partnering.  

Getting to the what (analytics) and the why (research)

The biggest advantage of combining behavioral data and research is that you get to explore and understand the what and the why behind a particular situation or phenomenon. Behavioral data will tell you what people are doing. You don’t need to rely on them to remember their actions – data is captured through interactions with the brand that you can assemble and analyze. Everything from purchase behavior to customer service calls to loyalty program participation can be tracked.

Once data is collected research can help with the why behind behaviors. It can prove or disprove hypotheses about behavior. It can uncover the underlying beliefs and attitudes that are driving that behavior, and identify opportunities to modify it. A holistic assessment is achieved when studied together. There may be information gaps and incorrect conclusions when studied separately.

An example of this is illustrated in the story below.

Detecting cholera in Rwanda

Nathan Eagle, now an American Technology executive and adjunct professor at Harvard University, set out to study how technology could be used to detect and maybe even predict cholera outbreaks in Rwanda. Using data acquired from mobile phones, he studied people’s commuting habits between and among villages. Patterns in his data began to appear: rather abruptly, movement would just cease. He hypothesized that some sort of disease or illness halted typical commuting behaviors.

Working with the Rwanda Ministry of Health, he compared actual flu/cholera outbreaks to what he was seeing in his data. Data showed that the changes in commuting patterns were correlated to the outbreaks. He used that to predict future outbreaks. After some time he discovered that his model wasn’t predicting a cholera outbreak; it was instead predicting when a village was experiencing a flood. He quickly determined that his initial approach had missed an important element: he had not collected any data, such as self-reported survey data, that would explain the why behind what he was seeing in the mobile phone data.

He modified his approach and developed tools via a mobile application to collect self-reported information from people on the ground as soon as he saw changes in commuting patterns. Without collecting the self-reported data and connecting those insights to his mobile phone data set he would have continued to misinterpret the cause of the commuting pattern changes and possibly have made inaccurate outbreak predictions. He now regularly incorporates the self-reported data when making predictions.

Opportunity and exploration 

Let’s look at a process for pairing behavioral analytics with primary research that can be used in a number of different situations. These situations range from solving brand, business and product issues to identifying new areas of opportunity and exploration for a company.  

Consider the business problem of a loyalty program that is not performing well. A common first action might be to go out and talk to loyalty members via focus groups because it’s easy and fast. All you need to do is schedule groups, have the facility recruit and in two weeks you’re talking to customers.

It’s not a bad place to start but there’s strong evidence that you’ll generate more compelling findings if you begin by understanding how customers are currently engaging with the loyalty program.

Here are seven tips on how to zipper behavioral data and research to address this bigger business question:

1. Start with what you know. Study how your customers are engaging with you today. Work with the data scientists on the analytics team to understand the attributes that are available to analyze.

2. Look for patterns of behavior. Ask yourself the following questions and use data to inform your current hypotheses about what is going on.

  • Are your super users (the most loyal) using your program less frequently than before?
  • Do you have super users with unused points or unredeemed awards? How many?
  • Are there patterns or different group members that are using your program less? What do their online shopping behaviors look like? What about their purchasing habits?
  • Are there groups of super users who are still using the program as they had before? What makes them different from the others?

3. Revisit and modify hypotheses. Through this analysis you’ve likely dispelled some myths and created new truths. It’s also likely that you will have new hypotheses; formulate and integrate them into your discussion guide.

4. Refine who you want to talk to in groups. From the behavioral analysis you’ll determine who you want to talk to in your focus group instead of relying on a broad array of loyalty members.

5. Survey the ones you’re with. Pull the sample for your focus groups from your database. It’s more efficient and less expensive to use your own sample, plus you don’t have to rely on an extensive set of criteria questions.

6. Quantify what you heard in groups. You may need to follow the focus group with quantitative research to substantiate what was learned in qualitative. Pull sample that allows you to mirror your reads in the groups but also plan for a representative read of your loyalty members.

7. Construct a final narrative that weaves it all together. Tell your story using both analytics and research. Deliver key insights and crisp recommendations for action as well as continued measurement.

Constant state of change

Customers are not one-dimensional – they make decisions rationally, emotionally and impulsively. Adding to the complexity, customers are in a constant state of acquiring, processing and synthesizing information, storing it for future recall – consciously or unconsciously. If customers are in a constant state of change, shouldn’t companies be more dynamic in how they understand them?

Behavioral data is the best place to start when looking to truly understand your customers, with research being a great tool for explaining that behavior. It’s not about choosing behavioral analytics or primary research – or saying that one is better than the other. The truth is that the two, when zippered together, can lead to unique insights that result in highly-tailored solutions to business problems.