Social Media Research: An Overview of Where We Are and Where We’re Going

Social media research is still a very young field in market research and, in fact, at this point we can’t definitively call it market research, for at least a couple of reasons (more on that in a second). It is not hard to see that social media monitoring has the potential to be a valuable component of most market research firms’ services, as well as a critical function in market research departments within companies. The reasons are obvious: the sheer volume of information that can be obtained, the ease with which it can be gathered, and the lower cost compared to much of traditional research methodologies. However, there is a flip side to this. With so much information, and the fact that the information isn’t answering any particular set of questions, it is very difficult to sort through and make sense of it all. In addition, a researcher has no demographic or other user-identifying information to work with, so they have no idea who is saying what. In other words, the information can’t be cross-tabbed into meaningful segments for the purpose of targeted marketing. These are issues that need to be sorted out in the future if social media research is to become a valid and important tool in the market research field. That said, social media research is poised to become a major part of research very soon, so those not getting on the plane now will be left at the terminal.

Following is an overview of where we are now and where we’re going in social media research. The article “Chatter Matters” in the Fall 2011 edition of Marketing Research was helpful in putting together some parts of this overview.

At this point in the fast-moving field of social media research, according to the Marketing Research article, we can identify three categories of social media research:
• Social media monitoring (or listening);
• Purposed communities – MROC’s and community panels
• Other forms, such as netnography and social media as a sample source, which are still evolving

We’ll look at two of the types of social media research listed above, social media monitoring and online community panels.

Social Media Monitoring
Social media monitoring, also commonly referred to as social media listening, is all about listening to what’s going on on the Web, from Facebook, Twitter, LinkedIn, blogs, and all the other social media hubs that are out there. Obviously, then, we’re talking about a huge number of conversations (millions every day) and a vast amount of information to sort through.

As touched on above, this mass of information poses a daunting task in making any kind of sense of it all. Even when narrowing down the parameters of the monitoring (i.e., when you’re monitoring just conversations about one or a few brands, or a handful of keywords, which is typically what a company would be doing), the amount of information can still be overwhelming. Fortunately, over the past few years, a multitude of services have popped up to help us as researchers make sense of all the conversations we’re trying to track. Many of these are free. There are also many fee-based services. Not surprisingly, the fee-based services generally provide a more comprehensive array of tools, offer more robust analysis, and have more accurate sentiment analysis algorithms. In short, you get what you pay for, so generally speaking, if you need comprehensive and robust tools, the fee-based services are the way to go. Some of the more well-known services include Radian6, Sysomos, eCairn and BrandWatch.

At this point in the evolution of social media monitoring, a number of important caveats need to be kept in mind when incorporating it into your market research programs. The top ones are: 1) Not all of the Web is accessible by monitoring bots/crawlers, such as large parts of Facebook; 2) Keyword analysis is still a work in progress in terms of accurately identifying keywords (though it’s quickly improving); 3) The ability to segment the information (e.g., by demographics, geographically, by business identifiers) is still not possible from most social media sources (though researchers are actively looking at ways to layer this onto the information); 4) The conversations we’re listening to don’t pertain to any specific question (as in a traditional survey) or in many cases even a particular topic. Thus, the context of much of the information is missing.

Despite these caveats, the magnitude of the information, and the ease and low cost of obtaining it make social media monitoring too attractive to ignore. Plus, in fairly short order the caveats/problems noted above should largely resolve themselves: keyword analysis accuracy is getting better; identifying customers, fans, key influencers, etc., is improving, through source tagging; engagement techniques with customers, fans, key influencers, can help in filling in the context gap by following up with them and asking questions and soliciting advice from them (and this not only provides valuable information, but also nurtures strong relationships with mavens for your brand).

In short, while we’re still in very early in the evolution of social media research, it is reasonably certain that it will eventually get to the point where it’s a fairly reliable, robust and comprehensive tool that will be a core component in corporate and organizational decision making in combination with traditional market research.

Following are the basic components of social media monitoring.

This is as basic as it sounds. It simply means counting the number of times a keyword comes up. This can be useful in areas such as the social sciences (how frequently a disease is mentioned), and in politics (how often a candidate’s name comes up), but the value is limited for business purposes because it doesn’t provide any insight as far as the sentiment behind the mentions.

Trends are simply looking at counts over time. For business applications, this is often done to see how campaigns are impacting the Web conversation. However, while trends can show direction, they don’t tell you the reasons for the trend.

Sentiment analysis
Counts and trends are fine to look at, but they are very limited in what they can tell you. What companies and organizations really need to look at is the qualitative nature of the posts or conversations taking place (i.e., are we getting more positive or negative comments about our company; what exactly are they talking about regarding our company), or sentiment analysis.

Because of the volume of posts or conversations, it is impossible to analyze them all manually (i.e., by a person or people looking through them), so one way to do it is to have a social media monitoring service apply automated techniques to do the analysis, using algorithms. Another method is to take a sampling of the posts/conversations and do it manually, though this can be quite labor-intensive, which is why many companies choose to have social media monitoring firms do this work for them.

Influence and identification
In addition to sentiment analysis, the other important component of social media monitoring is to be able to find out who the influencers are for the brand or organization. This is important because influencers are, by definition, influential in getting out positive (or sometimes negative) buzz about a brand/organization, and thus need to be engaged with and nurtured (or in the case of negative influencers, countered), so that their influence can be amplified. Engagement with influencers is a key part of any organization’s social media marketing strategy, so initial identification has to be a top priority.

Purposed Online Communities
Despite the value you can gain from social media monitoring, it is passive, and so doesn’t answer specific questions you need answered. Purposed online communities help to close some of the information gap inherent in social media monitoring. These communities are put together by the company or organization, are comprised of customers or members, and are asked to discuss specific topic areas, and/or asked to answer specific questions, so that the conversations are specific to a topic the company needs feedback on. Thus, these communities are more like traditional market research. The benefits of purposed online communities are that researchers can gather a large amount of information quickly, and it is fairly inexpensive once it is set up.

The primary downside of purposed online communities is that they are composed only of customers, and mostly enthusiastic customers at that, so you’re only getting feedback from a portion of your customer base, and nothing from dissatisfied customers or those who have left, as well as nothing from potential customers. Thus, you’re not hearing from the “silent majority” of customers, who may not feel as positive about the company as those on the panel. Just as important, you’re getting nothing from past customers, specifically those who left because they were dissatisfied, and so are missing this very important body of feedback and insights regarding how to improve products and services. Finally, you are not hearing from potential customers, so metrics like brand awareness, brand perception, and usage and perception of competitive brands can’t be gathered via this methodology. For these important metrics, researchers still need to use traditional market research methodologies.

Despite these issues, there is still value for companies that have the internal resources to start and maintain online community panels, as it helps them monitor the general sentiment of their customer base, screen new product/service concepts, solicit advice on how to improve products/services, and any of the other things researchers would use customer focus groups, expert panels and surveys for, but at a lower overall cost and with a much shorter turnaround time.

How Social Media Research Fits in with Traditional Market Research
Social media research, while promising, still needs a good deal of refinement and many enhancements before it can be an important and valid tool for market researchers. Currently, social media research is used by market researchers as a supplemental body of information to add to the traditional research they are doing. Right now, there are many areas of market research that cannot be replaced by social media research, and it may be quite awhile, if ever, before social media research replaces these areas of traditional market research.

Following are the most important elements of traditional market research that currently cannot be replaced by social media research methodologies:
• (Both social media monitoring & online community panels) Having representative sampling of the entire market (i.e., both customers and non-customers)
• (Social media monitoring) Knowing who your audience is
• (Social media monitoring) The ability to ask questions

Here are some of the types of traditional market research that social media research cannot now nor likely in the next few years replace:
• Attitude, awareness and usage (AAU) studies
• Lost customer assessments
• Brand positioning/competitive analysis

And here are some areas where social media research can currently supplement and enhance traditional market research:
• Advertising/message testing
• Provide general ongoing “pulse of the customer” information on company performance
• Provide quick, specific feedback on company performance in certain areas
• Provide quick feedback on new product/service ideas
• Provide ongoing information on the broad-based web conversation regarding your brand and products
o Correlate this conversation with current campaigns
• Provide ongoing information on competitive brands and products
o Correlate these conversations with known competitive campaigns
• Engage with brand mavens, to gain deep insights regarding company/product/service performance
• Gain broad, “macro” insights on general trends in the marketplace regarding product/service categories, the cultural/social landscape, and the economic landscape (both nationally and locally).

Given the above listing of what social media research can provide, we see the coupling of traditional market research and social media research as an exciting and revolutionary step in the field of market research. The potential benefits are immense for both practitioners and users of market research. As branders and researchers, we are looking forward to working in this emerging landscape and providing our clients more value by using this new combined, enhanced form of market research.