Tapping into social media for marketing purposes continues to top communications agendas at companies across the country. B2C and B2B marketers alike are very interested in exploring opportunities to “engage” buyers with their brand via blogs, on-line communities, social networking sites, Twitter and the like. Apart from the brand and (maybe) sales building capabilities of social media, marketers are also excited about the information and insights locked within blog posts, chats, and discussions that could help them identify trends, understand needs and wants, tap into opportunities, and make better, more profitable marketing decisions.
Now everyone seems to be in agreement—us included—that consumer feedback online could have important implications for a brand. The instantaneous nature and steady stream of information direct from customers and buyers in the category could indeed offer help to marketers looking to ensure what they are doing and offering with their brand is on track with what consumers expect, need, want, etc.
As former P&G exec turned marketing consultant Gordon Wade wrote recently in a piece for AdAge.com, “another rapidly growing source of data ripe for analysis is the public conversation on the internet.” He explains the availability of “new, second-generation” services that “enable new and more insightful ways to understand the online conversation.” PR exec and social media expert Peter Shankman maintains that, “Twitter is the ultimate focus group. I can post something and in a minute get feedback from 700 people around the world, giving me their real opinions.”
Their thoughts reflect the emerging conventional wisdom in marketing which places an increasingly heavy emphasis on listening to the raw voice of the consumer coming over the internet. Again the thinking goes that because it’s on-line and because there’s just so much of it, it’s somehow MORE representative of buyer opinion and predictive of purchase behavior. Wade went so far as to say “because we are talking about responses from thousands of people, this ‘qualitative’ data can suddenly take on the statistical validity of conventional survey data.”
Marketers still need to tread carefully when it comes to using information gleaned from on-line discussions ALONE to make decisions. The stories of the statistical validity of on-line opinion, you see, are greatly exaggerated.
In the language of statistics, a “valid” test measures what it purports to measure. Could engaging in a discussion about a brand on-line be a statistically valid measure of a consumer’s affinity for a brand? Yes, it certainly seems that it could be one. According to the Advertising Research Foundation’s Chief Research Officer Joel Rubinson, “there is published evidence from regression modeling that measures of brand affinity or equity can be calculated from social media, which, in turn, correlate with sales trends.” In other words, it could be a marketer might very well include questions about participating in public conversation on the internet in a survey as a valid measure of affinity. Yet this is NOT at all the same as saying you can or should look at blog chatter and twitter posts and generalize conclusions about the population at large.
Simply put, when it comes to meeting the requirement of a “valid” test of the opinions of buyers in the category, “public conversation on the internet” falls short. Yes, we’re (supposedly) talking about responses from thousands of people, but who are these people? Does a marketer know enough about the folks twittering away about their brand to conclude their opinions are at all representative of any the buying population at large? According to a recent Knowledge Networks study, just 1% of the total online population uses Twitter once a week or more. Does that doesn’t sound inherently representative?
Keep in mind that the folks tweeting could turn out to be an agent of the competition looking to steer its a brand in the wrong direction. Or, on the flip side, maybe it’s a friend of the brand manager trying to drum up positive chat about a promotion in order to help the brand manager look good to his or her superiors. You just don’t know who is expressing opinions or, very importantly, how they are related to buying behavior—if they are at all. As former American Marketing Association chairman Bill Neal once said of consumer-generated online feedback, “in many ways it combines the worst elements of non-scientific research—self-selection and advocacy—both positive and negative.”
Though online sources may offer feedback from more people than focus groups, the data is not necessarily any better at predicting real-world outcomes. Continuing in the words of Bill Neal:
I’ve seen too many brand managers observe a focus group or two and then make major changes in their marketing programs based on what they heard in the focus groups. Many of those changes are utter failures simply because the focus group was not a reflection of the real world. You have exactly the same problem with monitoring consumer generated conversation, except that now you have the equivalent of thousands of focus groups. Does this make the information more valid or reliable? Definitely not!
Consider the recent Kentucky Fried Chicken Oprah promotion to support the launch of its new grilled chicken. The chain had to cut short the free meal offer it initially extended via Oprah’s show when redemption exceeded projections. AdAge reported that online chatter “got nasty” when the promotion ended. According to Al DiGuido, CEO of Zeta Interactive, a firm that monitors blog chatter, “As soon as KFC decided to halt the promotion, their brand suffered a brutal backlash, plummeting down to 67% positive [on-line] buzz. With this overwhelmingly negative response, KFC did more damage to its brand by running an incomplete promotion than if they had just not launched the campaign in the first place.” Taking its cue from online reactions, AdAge reported that “KFC’s chicken run will probably end up one of the all-time blunders by advertisers.”
Did online chatter accurately fore tell the product’s doom? Doesn’t seem like it. In a follow-up story a few weeks later, AdAge reported that “early returns on Kentucky Grilled Chicken show heavy trial, strong repurchase intent and signs of a turnaround at the long-ailing brand.”
Just as in focus groups, the strong opinions of the few can still influence the thoughts of the many online. Suddenly everyone jumps on the bandwagon and it’s not necessarily indicative of a problem, trend, opportunity, or, very importantly, buyer behavior. Take the case of Tropicana when it rolled out a redesign of its packaging. Influential design and culture blogs started posting critical reviews of the new box design and others quickly jumped on the bandwagon with negative chatter. Complaining emails and phone calls to Pepsi, owners of Tropicana, ensued and the company decided to pull the plug on the new box. If design critics had embraced the design and written about how cool it was, the chatter train could have just as easily headed in a different direction and Pepsi might have stuck with the new design.
Tracking and mining social media, twitter, emails, and online sources of consumer feedback certainly has its place in the marketing research process. Understanding the language of customers, how they are using products, their reactions to different promotions, attitudes and perceptions of the brand as a lead in to further, more quantitative investigation is a very important, we might even say critical, thing to do. To quote Bill Neal yet again, “the shear volume of product mentions [in social media] and whether they are positive or negative is useful information and may (and I emphasize 'may') provide a signal that something is going wrong, or right, about the product and how it is being marketed.” Marketers just need to stay mindful of the limitations of the data harvested from online public conversations.
