Big data and data science help brands process information from inputs including social media, sensor data and online transactions, and they analyze this data at unprecedented speeds. Big data, both large and dynamic, is now used to drive decisions across every industry.
Big data has many compelling uses, such as fraud prevention and security intelligence gathering. However, the number one use case is understanding customer behavior. It provides data such as who bought what, when and where they purchased it.
Big data is used to analyze past behavior, identify patterns and then make predictions based on those patterns. It allows companies to identify new opportunities. For example, Netflix leveraged big data to create the hit show “House of Cards” by analyzing viewing habits of fans of the original UK version. Consumer packaged goods giant Procter & Gamble (P&G) leverages big data to tailor marketing campaigns to loyal customers and high-value prospects.
High-profile success stories with brands like Netflix and P&G have created an unquestioned faith in big data.
But if you use that exclusively, you may be missing the why behind your customers’ touchpoints. Thick data gives you this aspect through qualitative customer-centric data. Thick data completes the picture that big data starts.
Thick data enriches traditional analysis by adding the context that helps companies understand customer motivations and pain points. A major Mexican food manufacturer was trying to engage its customers more actively and generate more revenue. Big data and analytics indicated that a new loyalty software platform was the answer.
Before accepting this big data output, this client went to actual neighborhoods in Mexico City and outlying areas and conducted ethnographic research to understand the customers as people — what they valued, what motivated them, the underlying reasons for their behavior.
Through individual conversations and video diaries, the client discovered what the consumers craved was community, a sense of belonging, and that local influencers were the key to activating their loyalty. The company focused its efforts on identifying influencers in each local market and developing rewards that emphasized community. That tapped into the consumers’ natural sense of affinity and pride in their neighborhood.
Once this connection was established and reinforced, the loyalty system became useful and enabled scale. Ultimately, this increased sales from current customers by 40% and increased repurchase rates and share of wallet. Understanding consumers as people, not simply data inputs, made the difference, and this could not have been achieved simply by implementing new software.
A major European supermarket chain faced falling sales, and its big data analysis showed it was losing out to discount stores, in spite of the fact that customers still cared about quality. Unsatisfied, the chief marketing officer chartered an ethnographic study of the customers’ day-to-day lives.
Over two months, the researchers observed the customers shopping, planning, interacting with family and cooking. They discovered that customers’ lives were changing. Routines were disappearing, so meal planning was difficult. Family meals had disappeared, and families ate at different times and rarely at the dinner table.
Last, people were grocery shopping nine times per week, and had no loyalty to a specific grocer. They shopped wherever was most convenient. To realign with customers, this supermarket client changed the store experience to mirror its customers’ fragmented lives instead of lowering prices to compete with discount stores (Rasmussen and Hansen, 2015).
Traditionally, when a company wants to collect thick data, it needs to conduct a research study similar to the example above, sending a strategist into the field to interview, study and observe consumers. This is done over a few months or even a year. Once the study is completed, the collected research is analyzed and interventions are created to solve the business problem. The problem is that companies cannot afford to wait months to garner insights while customer behavior continues to evolve.
Another problem arises when follow up is needed with the target customer base. Typically, this means an entire new cycle of study, with a similar time and investment. Will it capture the same audience, and will that audience be committed to participating again? In a world where consumer tastes change rapidly, businesses need to make decisions in close to real-time.
In order to respond quickly, companies need insights that support this response capability: quick, agile insights to inform their decisions on a frequent, nearly continual basis.
A thick data platform, with a focused consumer segment, can provide these agile insights.
An example of an agile insights engine is The Motherboard, an online community of thousands of diverse and engaged moms, created by the Agile Insights Practice of WONGDOODY, a leading digital agency and an Infosys subsidiary.
Moms make 85% of purchase choices across all categories. These include food, clothes, appliances, cars, toys and life insurance. This is not a complete list, but it demonstrates that mothers do not buy only diapers and cleaning supplies. Making 85% of decisions, moms account for $2.4 trillion of consumer spending each year. Brands need to reach moms. Yet advertisements do not reflect the actual identities of moms.
77 percent of women do not relate to the ways they are portrayed in advertising.
Nor do they relate to the brand experiences being portrayed in advertising, such as the smiling mom cleaning the house or cooking. These statistics suggest that while brands have mountains of information on moms, something is missing.
Big data told brands that a typical U.S. mom is white and married, with 2.4 kids and some college education. This may have been true about moms of the past, but it is far from the modern mom’s identity. Many moms are breadwinners, nonwhite, millennial and single. Where are these experiences depicted in advertising? Where are the moms rushing out the door to work, and the single moms?
The Motherboard was created to gain insight about who the modern mom is and what drives her. It includes single moms, LGBTQ moms, immigrant moms, moms who stay at home, moms who work in corporate jobs and moms who work in the gig economy.
This community has been cultivated to provide feedback and input on brand experiences. The members are active and talk to one another. These moms also supply answers to the why and what if — thick data.
Thick data is pulled from surveys, one-on-one interviews, ethnographic research, video diaries and shop-alongs.
“If we see something interesting, we can reach back out and learn more. It’s an ongoing dialogue,” says Skyler Mattson, President of WONGDOODY. The Motherboard is valuable here because it is agile. Within 48 hours, a new idea can be launched on the platform, with quality responses coming back within 24 hours. This agility is due to a community of customers who are always on and engaged. Mattson explains:
“We are constantly engaging with them, so we aren’t just waiting for a brand assignment. Weekly, we are putting out topics of interest — asking these moms to weigh in on products or loyalty programs, or even things like travel planning, wealth management and decisions regarding health care and insurance.”
This is a far cry from the traditional approach of spending thousands of dollars to send a strategist into the field and getting a final report on your desk six months later. This is fast and flexible compared with the traditional ways of research. It is also continuous. This is not a “one and done” research study. “If we are getting mom input on a TV commercial, we adjust it, repost it and get feedback in real-time with each iteration,” says Mattson. It is a continuum and a community, not a one-time panel event.
The Motherboard insights engine is focused on moms, but the concept is applicable to any consumer or business segment. Using The Motherboard as a model, firms looking to use thick data to develop agile insights should keep its core principles in mind:
Bring together independent teams — made up of strategists, marketers and designers — to solve customer needs.
Flex and inform all stages, from strategists discovering insights to consumer input using the platform.
Developers conduct usability tests to inform the digital user experience.
Technologists test prototypes for feasibility.
Creatives learn about what messaging is best when bringing an innovation to market.
Inform and validate each of these stages continuously.
Research seeks to understand motivation, not just activity.
JustFab, an online-only monthly subscription retailer, is an expert in using big data. It has massive data — email addresses, digital ad reach, social media engagement and website traffic stats — from its customer relationship management system. It leverages big data to make important business decisions and serve its customers.
JustFab’s sales were going up, but sales were lagging within its customer base of mothers. While only one segment, moms make up 40% of JustFab’s customers. Big data did highlight to the retailer that moms are different from non-moms. JustFab’s moms are five years younger than non-moms, have a lower sale conversion rate than non-moms and buy more shoes and fewer clothes than non-moms.
Big data told the story that current JustFab apparel does not resonate with moms, and to boost sales, styles needed to change. Before advising JustFab to invest time and money in changing its products, WONGDOODY used The Motherboard to test this hypothesis. Using real moms, “we did a deep dive with one-on-one interviews,” Mattson says.
WONGDOODY contacted mothers all across the country to discover why they were not buying JustFab’s apparel. The discovery: Moms found the sizing inconsistent, and style was not mentioned. According to these moms, a shirt fit in one size, but did not fit when picked in a different style. And because they do not have time for returns, they decided not to buy products with inconsistent sizes. Moms did not buy JustFab’s clothes because they were worried about fit and the fact that they’d have to return those that did not fit.
Big data said the style of clothes needed to change, but thick data said the experience needed to change.
Instead of redesigning its entire clothing line, JustFab quickly executed two initiatives to build loyalty with its moms.
First, using its robust collection of product reviews, the retailer improved the user experience and personalized the product reviews that moms saw. JustFab showed moms reviews from women of similar proportions in order to increase their confidence in their purchases.
Continuing on the theme of fit, JustFab sent its customers tape measures so they could match their measurements to the size chart, which ensured that a specific item would fit. Without thick data, JustFab would have overhauled its entire clothing line, which would have been time-consuming and costly, and may not have increased sales sufficiently anyway.
Big data is an effective tool to identify patterns and make predictions. It helps companies understand what, when and how. It is necessary, but often not sufficient, because predictions should not be arbitrarily applied to unpredictable variables like personal motivations.
Since big data fails to explain the why of how consumers behave, thick data is a natural complement when dealing with the human element because it explains the why behind big data and what customers are doing.
Companies need both big and thick data, and an agile insights platform is the model to bring them together.
With customer tastes changing rapidly, an agile insights platform is a powerful ally for brands; it moves quickly, adjusts to changes in taste, is flexible and provides insights to act on with confidence.