Chapter 1

Organize for Listening and Define Objectives, Key Measures, and Conversations

This chapter outlines the first steps for doing listening research that were outlined in the Introduction:

• Organize for listening.
• Set objectives in relation to business goals.
• Define key performance indicators (KPIs).
• Determine the research subjects: the voices and conversation sources best suited to the listening program.

We will now discuss each of these in detail.

Organize for Listening

Because listening can be done by companies of all sizes, from small businesses to globe-circling enterprises, the ways that they organize to undertake the effort will vary depending on their goals, resources, and staffing. Organizational consultant Beth Kanter (2009) proposes three concepts for listening organizations that provide a helpful framework:

Whatever your company size or the organizational form you eventually adopt, Chris Bourdreaux, Converseon's organizational lead, tells us there are five principles to keep in mind for success (see Chapter 23 for details):

• Establish an internal champion.
• Begin with business objectives.
• Create a social media center of excellence.
• Focus investments on business processes.
• Manage culture and communications (listening changes companies and operations).

Is there a preferred model? Not yet. Companies are figuring out what is best for them using a combination of the organizational types and principles. As for staffing, some companies dedicate people to listening, others make it a shared responsibility. However, given the importance of social communication and listening to every company's future, firms will derive more enterprise value from figuring out how to bake listening into the business and its operations instead of merely treating it as a bolt-on or “cover the bases” requirement.

Set Objectives in Relation to Business Goals

According to Coca-Cola marketing expert Stan Sthanunathan, listening research needs to be anticipatory, innovative, action-oriented, and focused on making real business impact in order to make a valued contribution (for more on this, read Sthanunathan's essay in Chapter 21). He's talking about creating marketing advantage, not researching to explain. To achieve that outcome, we need to focus first on asking the right questions, questions that make a difference instead of just confirming what we know or validating past beliefs and/or experiences.

These questions can be either broad or narrow; each has its place and purpose. Some examples from cases that we reviewed are:

Broad Questions

Broad questions are wide-scope, often concerned with areas of general interest and driving trends.

• A global food company asked: “We're considering repositioning the company on the concept of ‘awesomeness,’ to stay relevant with younger consumers. What does this term mean to people? And if we use one or more of those meanings, will it risk alienating our current customer base?” (Chapter 13).
• A large CPG company asked: “How do people think about the economy now? What changes will they make to their behavior? Will any of those changes last, even after the economy recovers?” (Chapter 5).

Narrow Questions

Narrow questions usually deal with specific topics, categories, or brands.

• A financial services company asked: “What are the financial issues with which people are grappling nowadays, and how is this affecting their credit and credit card use?” (Chapter 8).
• A small business providing Internet services asked: “Why are customers dissatisfied with our services, and what would they like to see us doing instead?” (Chapter 13).

This quartet of questions reveals an important point: They are asked in different contexts and for different business purposes, which, from top to bottom, are:

• Rebranding
• Understanding mind-sets
• Developing message strategy
• Managing reputation

Don't fool yourself into thinking that questions like these are mostly for big companies to ask. The small services company out to understand how to restore its reputation faces problems that are nearly identical to Dell's during its “Dell Hell” period; the only difference between the two scenarios is company size. Take the time to phrase your questions properly, and if possible, collaborate with colleagues to bring in different perspectives. In my experience of leading hands-on listening workshops, participants are often surprised by how challenging it can be to make a question researchable. Hearing “We used 15 of the 20 minutes just trying to phrase the question!” is not uncommon. The better the research question, the better the research.

Define Key Performance Indicators (KPIs)

KPIs define and measure the progress of a business’ initiatives and goals (performance). Just to take social media marketing as an example, an exact search on Google for “social media KPIs” turns up about 29,000 results, with the top title being “35 Social Media KPIs to Help Measure Engagement (October 29, 2010).” The results list a variety of interactions such as downloads, reviews, sharing, and so on. Making the search more specific by including “listening” in the request (“social media listening KPIs”) rendered Google helpless, it couldn't find an exact match. Instead, it searched on the individual words which failed to garner very helpful results.

The crux of the matter is that listening insights in and of themselves are not revealed in standalone measures. Rather, they are often inputs to some other business process like customer service, R&D, marketing, sales, or public relations. For that reason, KPIs need to take the contribution listening makes to another process or function: their impact.

Companies that listen to social media conversations can compute metrics such as relating to business process improvements, which include quickening resolution time, increasing customer satisfaction levels, cost-efficiency gains, or profitability enhancement, see the JetBlue and Comcast cases (Chapter 14). Organizations also assess contributions to strategy by asking questions like these: How did listening insights frame our externally focused marketing, sales, or product strategies, for example, and influence the results they achieve? (See the Hennessy and Suzuki Hayabusa examples in Chapter 5.) They may also be concerned with response and cycle times: Did listening insights help us speed up time-to-market by improving our understanding of customers and prospects, and contribute to better coordination and teamwork? This isn't a complete list; it's merely meant to be suggestive. The KPI question can only be answered one company at a time, and needs to be asked at the front end, not the back: How will listening contribute to the work our company performs, how will we attribute its value, and what will success look like? Companies that are not able to answer those questions will never know or appreciate listening's effect on the business.

Determine the “Research Subjects”: The Conversations, Sources, and Voices Best Suited for the Listening Program

Once you have defined your scope, the next step is to locate the conversations and the places they occur, so that you can collect and, later, prepare for data processing and analysis. Where you listen, as well as the appropriateness of conversations collected, directly affect the quality of the research and, eventually, the insights you're able to derive. That said, accessing the “right” conversations requires a number of steps:

Choose Where to Listen: The Consumer and Brand Backyards

We know people converse both online and offline. Online sources command most of listening's attention for two major reasons: one, their variety—there are sites for every conceivable interest, taste, and enthusiasm; and two, for the relative ease with which they can be located and harvested. Specialized programs called bots, agents, or spiders are set up to collect conversations that truly occur everywhere. We place online sources into one of two groups: brand backyard and consumer backyard.

Depending on the project's aims, listening researchers use the backyards independently or mix the two in some fashion. For example, marketers of computer hardware, software, and services, like CDW (Chapter 10), concentrate on its brand backyard and professional forums, whereas a food manufacturer, like Kraft, explores consumer-oriented food and recipe sites, as well as its own sites (Chapter 11). In most listening initiatives, specialized software accesses the backyards and captures the conversations (see Chapter 2 for detailed explanations of the different types of software solutions).

A crucial point to remember: Wherever and whenever you listen, do it ethically, and harvest only publicly available or permission-based conversations. Respect the privacy policies of all the sources captured for the listening initiative.

Determine the Conversation Footprint

After deciding on the project's goals, it's time to locate the specific blogs, forums, social networks, and other sites where conversations of interest take place. You may just deal with a handful of selected sources if you are monitoring, as is often the case with customer service applications (see Chapter 14); however, if you're doing social research, it is essential to fish where the fish are. Defining the footprint is the fish finder.

This is a true research task. It's best to begin with known sources, like brand backyard ones, and then conduct searches to locate the places that are conversation-rich for your research goals. Keep in mind that you do not always need to include entire sites; only certain sections may be relevant. This is especially true for large Web sites, forums, and discussion boards, which have a variety of topics and attract different types of users or visitors. Set those that interest you aside, and get ready to take the next step: vetting the sources.

Vet and Select the Sources

The next thing you'll want to do is investigate the sources and choose which ones make the cut. Seth Grimes (2009), an authority on text mining, recommends four criteria for evaluating sources:

• Topicality: The source contains the information of interest, and enough of it, to justify including it.
• Focus: The source contains a high proportion of relevant information, and not a lot of extraneous information; said differently, the source has a high signal-to-noise ratio.
• Currency: The information is timely.
• Authority: The information is trustworthy.

The message here: Don't just use sources because they're popular (“We have to use Twitter”) or easily available, or included in the set of default sources provided by vendors. Employ them because they will contribute high-quality conversations for analysis.

Select the Right Voices

The two backyards contain an array of voices, and every site will contain its unique set. Quality research depends on your ability to discover and listen to voices that are especially relevant to your company, product, or service. Practically speaking, new projects should focus on ensuring that the main voices are captured, such as moms (Chapter 3), lapsed users (Chapter 7), or caregivers (Chapter 4) for example. Unless experience counsels otherwise, avoid being overly specific at the outset. Just as in traditional research, it's difficult to know in advance what all the segments of interest in products are; let them emerge through analysis.

Listening is bringing about changes in research thinking the marketing community is working its way through, which is worth mentioning. Conventionally trained researchers often challenge the representativeness of listening research, inquiring: Do the people holding the conversations (the voices) match the demographic profile of a product's users; were they randomly selected; and are they projectable to the target market? One reason they ask is because so much of the information on public sites is not personally identifying. Failing to ascertain the characteristics that are part of almost every traditional survey makes some researchers uncomfortable because the people conversing cannot be classified with certainty, which runs counter to historical market research practice.

Listening-oriented researchers counter that representative samples are less of an issue because “the traditional model of quantitative market research based on questions asked to random probability samples is fading away…and providing the opportunity for the first time in probably 70 years to really revisit the methodological foundations of market research” (Poynter 2010, also see R. Scott Evan's essay in Chapter 21 for related points). Nowadays, one of the most important goals is to ensure that the research reflects the people on a site or in a community who are interested in a topic, company, or product by “sharply defining the population you are going to study, and to get as much of that population as humanly possible…” (Webster 2010). Manila Austin and Julie Wittes Schlack (2010) expand on that idea:

If relevancy of insight is an important quality factor (which we believe it is), then researchers have more to gain by listening to the “right” group of people than they do by trying to generalize findings to a generic population. If you want to deepen customer loyalty, who better to engage than members of your brand's loyalty program? If your goal is to broaden your brand's appeal, then hearing from fans of your competitors’ brands may be the most useful approach.

Every company needs to answer the question of voices for itself and for each initiative, as different projects may have different goals. The tensions between tradition and innovation will be resolved and lead to wider adoption of a new listening-centered research paradigm.

Summary

Defining and setting up a listening initiative draw upon solid project management discipline: organizing, setting objectives, and establishing key performance indicators. Defining and vetting the conversation sources and selecting the voices to capture and analyze are vital to successful projects. Listening research represents new approaches to understanding consumers that are contributing to a new paradigm for market research.