There is something marketing managers seem to forget about the internet: it was made for people, not for companies and brands. As such, it offers managers a source of insight they never had — social listening.
Eavesdropping on consumers’ social-media chatter allows marketers to economically and regularly peer inside people’s lives as they are being lived, without introducing biases through direct interaction. Armed with traces of revealed opinions and behaviors, managers can at long last discover the manifestations and ripple effects of their actions on consumer behavior. Clear indications from marketing science underline how chatter affects sales, brand health, and even stock performance. Social listening competency will be critical to competitive advantage in the digital age.
But despite its potential, companies underleverage the social media stream for market intelligence. Analysts look for data confirming a predetermined viewpoint, or view the social media conversation as something to be managed rather than
to. They frame listening as a descriptive exercise rather than the high-potential strategic project it should become.
Some pay attention to social media data only when corporate crisis demands it. Although insights from social listening can and should drive corporate strategy and innovation, these are more likely trapped inside the marketing and service departments that “own” them. Social listening promises the Holy Grail in business: superior understanding of customers. Why, then, do managers fail to fully exploit it?
Econometricians, computer scientists, and information systems (IS) professionals often manage social listening efforts and their skills in database management and big data analytics are essential. But these hard scientists lack the social science skill set that allows managers to move from data to insight in the social listening world. At issue is the fundamental difference between information and meaning. True to their titles, IS professionals specialize in managing information. Their function is reductionist: bringing complex data down to the simple level of numbers — zeros and ones.
Anthropologists and the culturally sensitive analysts who think like them specialize in meaning management. Their function is to take complex bits of data and develop a higher-order sense of them. Information and meaning work at cross purposes. In managing meaning, context is everything while in managing information context is error and noise. When we give our social listening projects to information specialists, we lose an appreciation of context and with it the ability to extract the meanings that provide insight for our companies and brands.
Consider the tenet of “adequate sample sizes” and its antithesis online. With social listening data, one quote, one comment, or one posted picture can spark an idea with profound implications. A large pharmaceutical company, for example, learned about an unsuspected customer challenge through a single photo on Flickr. The image showed a man wrapping a part of his leg in foil after applying a pain relief ointment. It turned out that the medication left untreatable stains on certain fabrics, hence the protective foil. Executives had been unaware of the problem despite years of conventional consumer research. This single picture led to changes in the product and communications, and increased customer satisfaction.
The notion of “representative samples” that we impose when judging the value of quantitative research must similarly be put aside. Engaged social-media participants are no doubt a different breed from non-users, or those who read but do not contribute content themselves, yet they can serve a valuable signaling function nonetheless. Listening to internet chatter can provide a heads-up on which signals to take in, which are being amplified in the culture, and which need response from the firm. Many firms capitalize on this benefit but still listen only in an exploratory fashion, as a precursor to so-called “real research” that will determine the truth of what is being said. But theory development does not require representative samples. Non-representative consumers can be relevant because relevance depends on the question at hand.
Consider for example posters to an internet discussion group focused on video processing chips. Though this is a narrow group that is in no way “representative” of the broad sample of computer users, these heavily vested and deeply knowledgeable netizens provide critical knowledge about new product pick-up and quality concerns that is relevant to a broader population. In this case, social listening revealed how category-level loyalties can trump brand loyalties, reinforcing the need for first-to-market product strategies and a marketing engagement plan that includes presence not just in the branded community but also in general computer forums as well.
“Appreciating the qualitative” challenges notions about how knowledge advances. Quantitative reasoning serves a hypothesis-testing mindset or, at least, a quest for statistical relationships between known concepts. Social listening in its purest form does not presuppose anything and this unsolicited quality creates an opportunity to answer questions that managers do not even know they should ask.
A manufacturer of baby strollers, for example, operated for years on the assumption that it owned an acutely defined brand positioning, with “no nonsense” being one of the core brand associations. Qualitative analysis of thousands of online statements across several countries revealed a gap between the intended and realized brand perception. Not one statement reflected the core association, forcing managers to the realization that the no-nonsense positioning was lost on consumers. Their reaction was curiosity – an approach Isaac Asimov put this way: “The most exciting phrase in science—the one that heralds true scientific progress—is not, “Eureka!” (“I found it!”), but rather, “Hmm…that’s funny.’” Managers should drill into the data to ask questions, not confirm or reject hypotheses.
Meaning management also involves a deeper appreciation of social listening as a component of a broader meaning-making system, rather than as, simply, a data source to be exploited. Social listening data don’t stand alone: they are part of a complementary package of insights into consumers, consumption, markets, and cultures. Often managers stop at the first step of correlating what is “known” through company research and what is revealed on the internet. This correlation is often low, prompting managers to ignore or minimize the social listening data. This dismissive tendency is reminiscent of how managers treat focus group data: they love group discussions for their convenience and the comforting sense that they offer in-depth insight, but managers are quick to write off anomalous findings that do not align with their thinking. But, if social chatter reveals consumers’ lives in a way that commissioned research never can, then for this simple reason there are bound to be misalignments. Misalignments are a key source of customer insights because they challenge assumptions. Misalignments suggest a mandate to attend particularly closely to the data.
To get the most out of social media data, operations have to go beyond data scientists and the marketing departments that house them. Every executive has to be a listener. Chief executives at cutting edge companies actively listen to social media commentary—in real time each day, not through a sanitized monthly summary report from the marketing department. Senior managers across all departments likewise have to get their hands dirty.
Social media comments need to move beyond the marketing departments and service agencies that collect them and become part of monthly management committee meetings. Many tools exist to make it easy to integrate social listening data into reporting or other business processes — and this can be part of the problem.
For one client of Oxyme (the analytics firm author Rietveld co-founded) the team counted the number of dashboards available on market and customer behavior. There were 19. Instead of deploying dashboard 20, Oxyme created a daily email containing provocative positive and negative statements voiced that day by customers online. This simple approach yielded new consumer insights that would have been hard to uncover though conventional data-mining. For example, sentiment is always context dependent, and the analysis needs to be sensitive to what the affective feeling is about and who is doing the talking. Existing research on bad breath revealed the expected negative sentiment, but by analyzing the context, managers learned that the authors responsible for the negative comments were mothers and the solutions they sought to bad breath were for their children, not themselves. By focusing on the context of negative sentiment rather than its magnitude, the research revealed a new target audience. Now when managers look at their consumer sentiment charts, their interpretation is aided by the qualitative data that provide context and meaning.
To leverage social media for customer insight move beyond the science of data management to the art of interpretation, embrace the context offered in qualitative commentaries, and don’t delegate social listening to the marketing department.