Predict the Future; Don't Explain the Past

Stan Sthanunathan, Vice President, Marketing Strategy and Insights, Coca-Cola

When CEOs want to know how a product launch went awry, how purchase habits have changed, or where the company is gaining ground, they ask research for answers. But what if they ask about what is going to happen instead of what already did? Then researchers might be stumped. Yet if we want to increase the impact of market research, we need to focus far more on getting organizations to predict, shape, and capitalize on change than on carefully explaining their past. Our role needs to shift to helping companies dream about and create a more compelling and rewarding future, and driving the transformational change that will get them there.

Instead of creating the newest, greatest research tool, which will quickly become irrelevant and outdated, both agency and client-side researchers must engage in a change of mind-set that puts the ultimate needs of the business front and center. A contemporary research approach should revolve around the four key elements described here.

1. Anticipate New Opportunities

• Inspire
• Transform
• Seek what we don't know and how to figure it out

If a researcher had gone to consumers in 1886 and asked them what they would most like to drink, not one of them would have asked for something loaded with bubbles and a strange, sweet, syrupy taste. But Dr. Pemberton didn't ask what people wanted then; he anticipated what they could want and carried a jug of his unique mixture down the street to Jacob's Pharmacy to be sold as Coca-Cola, for 5 cents a glass.

Similarly, if you asked people in the late 1850s how they would like to get from one place to another, there would not have been a statistically significant response for developing the automobile. Since one did not exist yet, no one knew they wanted one yet.

The pace of change in our society today and the diversity of its inhabitants make it ever more critical for market research professionals to act as anticipators, and even agents of change. We need to get closer to clients and consumers to find out what is keeping people up at night, and then tailor products and services to meet those needs. The goal is to shape change, not follow it. And we need to create the change people have not yet dreamed of.

2. Use Innovative Approaches

• Observe
• Listen
• Synthesize
• Deduce

Getting ahead of societal change rests on finely honed powers of observing and listening to people. It also relies on the ability to synthesize what we learn and to draw logical deductions about the kinds of products and service-delivery methods that might better meet their needs. We do this by tuning in to people with an open mind and waiting to see how they illuminate us. We start the process, as the ARF makes clear, by “listening for the unexpected,” rather than quantifying the expected.

Thanks to techniques such as BlackBerry ethnographies and the practice of “netnography,” there are many ways to “listen in” to people's hearts, minds, and lives today, which never existed before. We can develop ways to sweep the content of the Web and conduct advanced analysis on it to really learn about people's passions, dreams, and what is keeping them awake at night.

If you focus on the well-accepted adage of leveraging what you already know, you run the very real risk that everyone else in your industry already possesses that same knowledge—and is acting on it. There's more power in figuring out how to know what we don't know.

3. Turn Insights into Action—Now

• Capture info
• Organize insights
• Create impact

It is possible nowadays to calculate results immediately with real-time feedback from consumers. These responses drive an immediate reaction in terms of tailored marketing messages and even product development. We can use technology-enabled observation to stay ahead of the game.

Coke recently used social media to engage people in creating a new flavor of vitaminwater, right down to its packaging (see Chapter 6 for a case study discussion). A process that normally takes two years and millions of dollars of investment in formal market research and trials was turned on its head, and conducted in three months for a few thousand dollars, by harnessing the power of the crowd to co-create a new product.

Organizations are also now using responsive outdoor advertising, with messages that change based on the time of day. Even more sophisticated versions can quickly scan passers-by for key demographic data, such as age and gender, and then display an advertising message targeted to that demographic, or instantly transmit a coupon for the product to a mobile phone.

4. Focus on the Business Impact

• Connect results to business growth and health
• Demonstrate the connection to senior leaders
• Provide inputs for strategy

As research professionals, we are appropriately concerned with the quality of our methodologies and reliability of the insights we gather. That vigilance is critical in terms of our own professional ethics and in the trust others put into the new ideas we provoke. But we also need to understand that the “back room” machinations of good research are not what earns a seat at the decision-making table or the ability to influence corporate direction. In other words, a pristine research project is meaningless unless it shapes business and drives value for customers. We must always draw a straight line between what we're doing and learning and how it will impact our company's growth and health. We also need to be able to demonstrate that connection to senior leadership.

If we direct our focus on ROI and value creation, we'll be key players in helping the company figure out how to execute better in the marketplace and increase value in the business.

Making It Happen

Of course, there is a requisite skill set to conducting quality research, and there are new techniques to be learned almost daily. We also need to master new mind-sets (see Chapter 4) and be willing to experiment a little and get out of our comfort zones.

Just a little open-mindedness can compel us to embrace new opportunities for learning what's important to consumers—even when they can't completely articulate it yet themselves. And technology is making it possible for us to achieve the paradox of customizing our messages and products on a massive scale.

At the heart of doing market research right in today's marketplace lies a willingness to stretch out your neck and try to figure out where change is going to appear next. It lies in listening to people's thoughts and desires to learn what really moves them. Armed with that inspiration, you may be able to predict which products and services will help them meet those overwhelming needs.

The Future of Marketing Research and the Professional Researcher in a Digital World

Artie Bulgrin, Senior Vice President, Research and Analytics, ESPN

Irish philosopher and management specialist Charles Handy once said that “measuring more is easy; measuring better is hard.” This has always been the case, but never more so than today. In fact, marketing or media research is often misunderstood and undervalued, due to its complexity, cost, and perceived limitations. Exacerbating that situation is the growing illusion that consumer measurement has become easier and even more robust, due to the instant gratification of data that flows constantly in a digital economy.

But while the quantity of data in today's world is apparent, the quality and clarity is not. Looking ahead, I strongly caution that this is most certainly an illusion. I believe that the near-instant availability of data—as a by-product of our digital transactions—will raise more challenges for marketers in the future, and will require us to rely increasingly on trained research practitioners to make sense of it all. My fear is that the ubiquity of transactional “analytics” will continue to lead some organizations to reduce any investment in or reliance on the traditional marketing research process and classically trained research professionals—an approach that is both shortsighted and risky.

At ESPN, I am very fortunate to work in an organization that generally understands and appreciates the value of quality research. However, we didn't establish this rapport overnight. Through collaboration, thoughtful communication, and strategic focus over time, my staff has been able to gain the trust and respect of our management team. But, as our business has evolved in recent years, especially in digital media, there has been an inclination toward democratization and decentralization of data throughout the organization. Though inevitable—and even practical—this trend inherently carries risks that the data's end users may not understand a dataset's limitations or operational definition, and what it represents, which could lead to misinformed decisions. Nevertheless, as these end users become more comfortable and independent using internal analytics—complemented at times with “free” research downloaded from the Web—there is the threat that the role of the traditional marketing research function may be compromised, moving us from data democratization to data anarchy. As a result, we increasingly find ourselves defending our own findings, while disputing conflicting claims from other sources or conclusions derived from our own internal analytics.

At times like these, it's important to revisit what the role of marketing research is to an organization. In 2004, the American Marketing Association revised its definition of “marketing research” to the following:

Marketing research is the function that links the consumer, customer, and public to the marketer through information—information used to identify and define marketing opportunities and problems; generate, refine, and evaluate marketing actions; monitor marketing performance; and improve understanding of marketing as a process. Marketing research specifies the information required to address these issues, designs the method for collecting information, manages and implements the data collection process, analyzes the results, and communicates the findings and their implications.

The key word in this definition is “information,” making it clear that the marketing research professional's essential role must continue to encompass managing the process that ultimately converts data into tangible and actionable information. This starts with ensuring the reliability and validity of the data that flows into the process. Whether it is internal analytics or a DIY online survey, there is no substitute for the traditional marketing research function to protect this integrity and manage the process. Though it certainly should not be replaced by self-serve knowledge solutions, that's my fear, as transactional, social, and behavioral data become even more inexpensive and accessible.

To me, research transformation means that analytics need to be considered part of the marketing research process to understand the consumer. While listening is also part of this transformation, it should not be limited to online or digital sources, either. In fact, this book makes it clear that, when not filtered properly or placed in proper context, there are concerns over the quality and representativeness of online sources. This affects the validity and reliability of the data. The author is also careful to point out that while the volume of branded conversations or buzz on the Web is well known, it actually represents a small percentage of all word of mouth. This is a great example of a digital knowledge gap which, in this case, is carefully filled today by research conducted by the Keller Fay Group.

Yet despite this available knowledge, so many organizations still rely solely on online sources. Why? Because it's easier and faster. Overcoming this challenge requires a professional who is able to communicate and demonstrate the value of the added cost and effort necessary to fill the knowledge gap. The benefit is broader vision and greater clarity.

At ESPN, we integrated the business analytics function with the marketing and media research functions two years ago. This allowed us to not only better understand the consumer, but also to improve and manage the quality of our internal data. The idea was to enhance the reliability of our internal analytic data and augment it with secondary and primary consumer data—including listening—to provide greater context, insight, and information. It has worked very well so far. But I know that maintaining this role in coming years will be a challenge, as it will be for all of us in this industry. Measuring better will continue to be hard work!

From Narrative to Insight: Trends in Social Media Intelligence Platforms

R. Scott Evans, PhD, Director, Social Media Thought Leader, Microsoft

Social media is only one of many sources for deriving business insight. But in order for it to become a primary contender for research budgets and business leader mindshare, it must advance on six fronts:

As social media analysis increasingly finds a place in corporate research centers, methodologists will challenge the validity of social-media-derived insights. Social media monitoring platforms suffer serious shortcomings when it comes to thematic mapping, sentiment measurement, and language capabilities. In comparison to conventional survey research, there are serious concerns about the diversity of the sample. As for influence structures and data integration, these are far down the wish list of most platform developers. However, researchers will argue that these two capabilities will greatly contribute to both the prescriptive utility and overall validity of the analysis.

While my comments on the state of social media appear pessimistic, I am actually optimistic on this front, given some recent trends among platform developers and companies that seem to want to better understand the insight social media has to offer.

Mapping Conversations and Measuring Sentiment

The first important development impacting social media analysis is the rapid convergence of listening platforms with the more mature text analytics industry. Though accessing content is an ongoing dynamic enterprise, the fundamentals of the technological solution are in place.

Emerging text analytics hybrids, built on natural language processing (NLP), are offering researchers sophisticated tools for building detailed multitiered classifications. This makes prescriptive insights both intuitive and defensible, by being able to break down issues into their constituent parts and matching themes to changes in volume and sentiment.

NLP offers valuable advances for improving sentiment measurement. Capturing word relationships at the clause level lets emotive words be more readily associated with the appropriate subject. Comprehending sentiment at this level is a key building block for measuring disposition at the level of individual themes. Without an accurate gauge of emotion, it is difficult to translate themes and volumes into business insights.

The Language Conundrum

While language is of lesser importance for regional businesses, it is critical for global companies. Understanding what social media can tell us about a message's impact in a global campaign can be a critical part of the evaluation process. Initiatives that focus on digital media must make social media insight a central part of planning and eventual evaluation. Without the requisite coverage, social media must take a back seat to more expensive conventional tracking surveys.

Fortunately, machine translation continues to improve. Both social media platforms and their text analytics platforms are continuing to experiment with automating the translation process. While success is mixed, there have been constant improvements, which suggest that a “good enough” solution may be on the horizon.

Influence and Optimization

Unfortunately, the multiflow nature of online influence means that considerable work still needs to be done to capture the way in which memes and meaning are generated and propagated throughout the Web. Memes are ideas and messages in the social media environment that encompass forms of expression that transcend written language. Videos, audio, images, and even certain types of online behavior are forms of memes that can spread rapidly across platforms. For example, simple Web analytics and mapping hyperlinks or followers do not address who should be targeted, and with what message. The road map may be in place, but the tools for understanding the actual traffic flow and the relevance of the goods being transported are currently inadequate.

From a tactical vantage, it is essential to know who is carrying what, and to whom. For example, the PR firm that needs to assess optimal communication solutions must combine the “who” with an understanding of how messages transform and cross platforms, as well as the role specific agents play in this process. It is absolutely necessary to have all the other elements in place. Exhaustive mapping and accurate sentiment are vital components for researchers to trace the permutations of memes across all social media channels. Aligning actual connections and content are crucial for building actionable influence analysis.

Validity and Representativeness

The great advantage of social media is its historical record. The transition from telephone to online panels required extensive simultaneous surveys to calibrate the online medium with its conventional counterpart. This process can be replaced in part with time-series. Building benchmarks and trends for themes and sentiment enable researchers to compare results with both known events and already-validated datastreams from other sources. As these comparisons progress, we will begin to understand both the strengths and weaknesses inherent in social media as a means to generalize market trends.

Those who argue that social media analysis today focuses on unrepresentative voices and public expressions fail to recognize that such positions become irrelevant if social media proves to be a market bellwether. Critics who cite the requirement to cover private conversations—or the need to traverse the entire social media universe—fail to consider the implication of matching time-series events. Those who refer to the many gaps in the crawling profiles of most social media monitoring platforms do not recognize the inherent sampling limitations and weighting schemes of most inherent methodologies. Regardless of the nonrandom sampling of the social media universe, consistent matching of historic patterns does give credence to the potential validity of social media analysis.

Insight Integration

The goal of social media analysis for business is simple: to improve our understanding of the market. This means that we must begin to integrate what social media offers with the numerous other sources of insight currently available to decision makers. Since social media represents a valid voice of customers and market influencers, you disregard its relevance at your peril. To the extent that social media analysis matures along the lines identified in this essay, it will increasingly become an important source of insight in decision-making circles.

Synthesize, Clarify, and Inspire Action: Keys to Effectively Communicating Listening Insights

Manila Austin, PhD, Director of Research, Communispace Corporation, and Britta C. Ware, Vice President, Research Solutions, Meredith Corporation

What is the future of listening? When we were asked to write an essay to explore this question, we put our heads together and wished we had a handy crystal ball—or even a geeky market research magic 8 ball—that could provide a ready answer. We don't have any such object, of course, but we do have a shared passion and point of view.

Along with countless others, we have first observed the remarkable shift in marketing and advertising from persuasion to engagement. We've witnessed the commensurate shift in market research from studying “respondents” to listening to consumers. Second, we have experimented with and embraced listening technologies as they have emerged, championing the voice of the consumer in our work. And, third, we have taken a step back to look at the state of listening today, and have found many examples of how brands are staying relevant, and leading, as a result of listening to real people in their own words.

But we also see room to grow. In our respective roles, we struggle day to day with a common problem, namely: how to manage the complexity that listening creates. We attend conferences and see a lot of data points; we read reports that faithfully describe findings…and more findings…and more findings; and we participate in the industry debate about how to transform market research, listen to and engage consumers, and maintain data quality.

Listening creates complexity on two fronts: It challenges many of the assumptions underlying traditional market understandings, and it generates a tremendous amount of information. If market research is going to evolve with the times, the future of listening will have to be about more than simply listening, in and of itself. While the act is necessary, and even critical, ultimately, it is not enough. Communicating what we hear in a way that synthesizes data points, clarifies insights, and inspires action is the second half of the equation. We need to do more than listen to the consumer's voice; we need to bring that consumer to life.

Listening Creates Complexity

In the past decade, we have seen an explosion of online listening technologies. They are social, engaging real people in conversations and co-creation; they are nearly infinite in scale, incorporating a multitude of voices from nearly every walk of life; and they are evolving rapidly. Thus, the possibilities for every kind of listening—from passive eavesdropping to co-investigation and collaboration—can seem endless. While these developments create great opportunity for market research, the use of social-media-driven technology fuels the quality debate that's been raging for years. It fuels a sense of concern about declining response rates, questionable respondents, and sample representativeness. Listening, particularly to hundreds or thousands of people in a conversational way, blurs the boundary between quantitative and qualitative, quasi-experimental, and humanistic methods. This can be confusing, and can further complicate how we interpret what we learn, while undermining confidence in the insights we uncover.

Therefore, one factor that must certainly play a role in the future of listening is our ability to reconcile this conflict. We need new ways to understand and explain what quality looks like as we expand our repertoire to incorporate consumer-led versus researcher-led methods. It will fall to us as researchers to make informed trade-offs that allow us to continue to listen through technologies as they develop, and to educate colleagues and clients about what we are risking—and gaining—by shifting our focus and methods.

Rather than doggedly applying familiar standards to the rich, unstructured information generated through naturalistic online conversations, we find it is often more helpful to focus on what is actionable (as opposed to irrefutable). This means using a range of techniques to pragmatically generate and test hypotheses. Integrating elements of both humanistic and experimental approaches allows us to produce timely, “good enough” research targeted to specific business needs. It also requires that we generate a broader definition of quality and consider how online, social, consumer-led research can actually strengthen validity.

For example, we see quality enhanced by trading artificial for naturalistic settings. One of the great benefits of social media is that it brings research into the consumer's world and allows for comfortable, convenient participation. Creating an intimate, natural space for consumers to relate to each other and generate their own discussions yields authentic and detailed insights that are likely to be missing when we rely solely on quasi-experimental surveys. Our experience has also shown us that when we trade anonymity and distance for relationship, we do a better job of engaging and motivating consumers. Simply said, people are more forthcoming, and their input is more useful, when they know whom they're talking to. A common fear is that feedback from people will be inflated and biased when they engage in long-term, branded research. We have actually found that the opposite is true: Ongoing relationships result in higher-quality, more genuine feedback. When consumers care about a brand, they are more likely to provide useful insight—even when it's unfavorable—precisely because they feel more invested in the brand's success.

Listening Requires Synthesizing, Simplifying, and Clarifying What We Hear

The future of listening will entail evolving and clarifying the process of research; it will also require us to clarify the product of research. Listening to the customer's voice is potentially disruptive, in that it tends to provoke as many questions as answers. It is essentially generative in nature, which means it can produce a great wealth of rich information—more voices, more perspectives, and, ultimately, more specific and actionable insights.

But the volume and diversity of data generated through listening can be overwhelming. We live in an era of shrinking attention spans, where information must be transmitted and consumed in 140 character snippets. Therefore, the richness produced by listening cannot afford to be reported in all of its complexity and detail. Somehow, we need to streamline the outputs of listening so that they are digestible and linked to business objectives.

This task is challenging precisely because—as Chip and Dan Heath explain in their bestselling book Made to Stick: Why Some Ideas Survive and Others Die—our awareness of all that detail and complexity can actually get in the way of our ability to effectively communicate what we know. Powerful ideas are both simple and creatively crafted; they reduce intricacy into pithy and compelling statements that allow others to absorb information and pass it along to others.

Five Keys to Effectively Communicating Insights

While there are likely many ways we are all tackling this problem, here are a few tactics that have helped us synthesize, clarify, and communicate insights about the consumer:

The practice of listening will require all of us to elevate and evolve how we apply our expertise, how we communicate insight, and how we interact with business partners to help solve problems. It is true, of course, that many of us in market research are already doing these things. However, as the art and science of listening continues to gain traction and move into the mainstream, these qualities will shift from nice-to-haves to must-haves.