In recent years, “big data” has become a popular term, especially among marketers. In short, big data refers to challenges presented by the exponential growth and availability of data. While big data’s most defining characteristic is obviously its large volume, simply amassing a huge trove of data is worth little without the ability to effectively harness it, conduct the appropriate analysis, and ultimately provide reliable answers to important questions.
While some may view big data as a threat to traditional market research, we view it as a complementary source of information. Big data is great at answering the “what” questions. For example, it can be an effective tool for analyzing sales figures and uncovering trends and patterns. However, it often falls short in being able to answer the “why”. Social media listening, which falls into the realm of big data, might answer some of these questions, but a focused survey is usually more effective and efficient for answering a specific question. As our founder, Charles Colby, quipped, “Scraping from social media instead of using surveys is like being lost and looking for landmarks instead of just asking directions.”
This is not to infer that big data is less valuable than market research, just different. A key part of what we do as researchers is identifying the tools and methodologies that provide the best fit for the information needs and objectives of our clients. We have found that using big data in conjunction with traditional market research methodologies is a powerful combination, and has enabled us to deliver deeper, more reliable insights than either information source may have been able to provide alone. As illustrated in the case studies below, the combination of big data and traditional market research has a variety of applications, from strategic customer segmentations, to customer satisfaction and loyalty, to predictive modeling.
Case Study #1: Using Customer Segmentation to Guide an E-service’s Differentiated Marketing and Product Development Strategy
An online service that provides subscription-based information services to millions of consumers was interested in profiling its core subscriber base and prospective customers to support development of a tailored marketing strategy and support product development activities. To address the client’s goals, Rockbridge designed an online customer segmentation survey that included motivations, barriers and needs, indicators of customer value and targeting variables. The analysis included both the survey data and behavioral data from the client’s data warehouse. The analysis ensured accurate prediction of the segments using only subscriber database elements to allow the company to score its entire customer base without collecting additional survey data. In contrast to a pure database segmentation, the use of survey questions in the development of the segmentation ensured an understanding of motives and product needs to guide a differentiated marketing and product development strategy.
Case Study #2: Linking Customer Satisfaction and Loyalty of an Online Retailer to Their Bottom Line
Rockbridge conducts an ongoing satisfaction and loyalty study for a large e-commerce client. To understand how the performance areas being measured impacted the bottom line, we conducted a linkage analysis between customer loyalty, key areas of improvement (from the survey) and transactional data provided by the client. The results revealed a direct linkage between the loyalty index, and actual spending and repurchasing behavior. Moreover, Rockbridge was able to quantify the dollar amount of return for improvement in key areas, enabling the client to effectively prioritize and allocate resources.
Case Study #3: Using Statistical Modeling to Predict Conference Attendance for a Professional Association
A leading professional association was interested in understanding the factors that drive intentions of attendees to return to their Annual Conference the following year. To achieve this objective, Rockbridge created a statistical model to predict the intent to return using questionnaire variables. The resulting model shed light on conference elements and attendee characteristics that explain and predict intentions. The results were provided in a spreadsheet that allowed management to better understand attendance dynamics through interaction.
Big data and survey research methods can clearly be valuable allies in approaching many different problems. However, the common thread among the examples above, and the foundation of how we approach big data, is taking small bites. Our focus is not on simply using more data, but on leveraging proven market research methodologies to use the data our clients already have more efficiently – precisely targeting the information critical to their success.
Written by: Robert Devall, Senior Director, Client Services