Brands have always wanted to know more about consumers and how to turn them into customers. What makes them tick? Why do they buy some products and ignore others? Why does a product feature that didn’t seem all that remarkable turn into a runaway hit?
Before around 2007, companies relied on focus groups, consumer surveys, and aggregated demographic and psychographic data to answer these and other questions about consumer preferences and opinions. Although all these methods are still in use today, the “firehose” of online data puts modern consumer insights at the fingertips of every company — and, best yet, available in real time. From social data and online reviews to news and proprietary data sources, more data is available today and it’s much easier to analyze. There are even plenty of free tools to get you started.
A Brief Look Back
In 2015, Adidas launched a robot-powered, on-demand sneaker factory. Why? To respond faster to consumer needs (or maybe whims) while cutting manufacturing costs. Many decades prior to this, Henry Ford commented that customers could have their Model T in “any color as long it is black.” He felt that feedback from his sales team about the need for more colors was unnecessarily elevating the desires of a very small percent of the market.
In the 1980s, Wells Fargo was the first bank to offer 24-hr. customer service by phone — and not with interactive voice response but living, breathing agents. Yet this was not something the bank did in response to customer demand. In fact, 24-customer phone service barely made the list of desirable features in one of the bank’s consumer surveys. Nobody cared. No, the company did it because it could.
A phone center consolidation and higher-capacity trunk lines made it possible to easily and cheaply route calls between call centers regardless of where they originated. Nevertheless, customer response was immediate and extremely positive. It took the bank’s major competitors six months to catch up. When asked about this, Wells Fargo’s then head of retail banking Bill Zuendt noted that asking customers what they want isn’t always the best way to source ideas for new products or product features.
So, what does all this mean? Do consumers know what they want? When they tell you, should you instantly react? How can you separate the signal from the noise?
Social Media: The Not-So-New Secret Sauce for Consumer Insights
General Motors set up the first full-time consumer research department in the auto industry. It was a leader in market-based decision making — sending out three million mailings a year to customers and non-customers. This was back in the 1930s!
Today, GM has a social media center of excellence (CoE) into which it has poured substantial time, money, and executive leadership to leverage the value of consumer insights. The CoE unites the strategic and tactical social media efforts of GM’s marketing, communications, and customer care teams. Although the center’s mission includes crisis management, it also uses consumer insights to quantify conversations around specific issues, which helps various departments manage brand perception at a granular level.
Online Data — The 24 x 7 x 365 Focus Group
Focus groups have been around for a long time. Sociologist Robert Merton came up with the idea in the 1940s. Merton was asked to find out how Americans were affected by mass communication, especially war propaganda, during World War II. After the war, focus groups were quickly adopted by businesses. In the 50s, focus groups helped Chrysler boost sales of their convertible by changing their advertising to appeal to women.
Focus groups have not disappeared. But as blogger Zuzanna Pasierbinska-Wilson points out, “The emergence of social media now provides brands with unique insight into the minds (and buying habits) of consumers. Brands get an unfettered view of consumer opinion about their products and even competitor offerings by monitoring and analyzing social network data (“social analytics”). Think of it as the new focus group without those paneled walls, water glasses and two-way mirrors.”
Another traditional method for learning how customers or prospects think is the survey. And, of course, they’re still in use — often in online form. Yet, surveys regardless of form have several drawbacks. One is time. A comprehensive, well-constructed survey takes time and research to create. And, even more important, for respondents to complete. So, they’re often abandoned.
And, as Netflix discovered when it surveyed subscribers about movie and television likes and dislikes, what people said they liked to watch and what they did watch were often quite different. Sometimes our survey responses are more aspirational than truthful. But when a positive or negative Tweet storm kicks up over a product or service, you’re much closer to the unvarnished truth.
So, there is undoubtedly a ton of social media data — qualitative and quantitative — to work with and from which to harvest consumer insights. But how do you handle it all?
Artificial Intelligence — Customer Insights Heavy Lifting
People are pretty good at understanding language. We know how to interpret words based on context, tone, and what we know about the person saying them. And this applies to the words we read on social media. We know when someone is being sarcastic or making a joke. We understand puns, figures of speech, and colloquialisms. We know, for example, that a milkshake duck is not a duck that drinks milkshakes.
Milkshake duck is an Internet meme, which describes phenomena that are initially perceived as positive, only to be revealed as deeply flawed.
While computers have gotten — and continue to get — a lot better at interpreting those things, a fully automated tool can’t compete with a human when it comes to understanding and categorizing the language used in social posts.
On the other hand, no individual or team can manually analyze a trillion tweets or Facebook posts, or the content of a thousand online forums. That’s where AI and machine learning comes in. For extracting customer insights from social media data, a machine learning algorithm can be trained to categorize posts or look for patterns in text just like you would. Only a lot faster.
New Kid on the Block — Image Analytics
Until recently, extracting consumer insights from social media was primarily a text-based endeavor. Social analytics tools looked for patterns and signifiers in text from which it was possible to measure more abstract concepts like sentiment.
But as social media has become more visual, with millions of GIFs, still images and videos, organizations are starting to look at image analytics as a way to “unpack” a deeper layer of meaning that can add to the insights gleaned from text.
For instance, many people take pictures of food. Food they want to eat. Food they are eating. Food they’ve prepared.
Yogurt is a good example. A yogurt brand could glean a lot of insight from the pictures people take of the yogurt they’re eating. What’s the context — at home, at a café, on a park bench? Is it one person or a group? More than one yogurt in the shot, or several? Are things being added to the yogurt? What kind of things? Healthy? Or guilty pleasures?
It’s easy to see how these images and their context could hold a lot of information that yogurt marketers might use to understand their customers, and yogurt lovers in general, better. You would talk to fitness enthusiast yogurt fans differently than the guilty pleasure group. That’s why image analytics have already gained a secure foothold in lifestyle marketing.
As this discipline matures and the software improves, image analytics will make significant contributions to the pursuit of consumer insights.
Get social insights delivered to your inbox.
Consumer Insights Can Help Your Business Right Now
Every day more and more businesses are using social media analytics to discover and harness the power of consumer insights. They’re using consumer insights to explore audience segments, refine personas, track customer response to campaigns, identify competitive threats, and much more.
To learn more about the value of consumer data powered by AI technology, download our business insights guide today.