The good word from devoted customers may not always be the most effective promotional tool.

In recent years, firms have turned to non-traditional marketing campaigns to generate buzz about their products and services. Indeed, positive word-of-mouth is anecdotally cited as the secret behind such successes as Chrysler’s PT Cruiser and the revival of Pabst Blue Ribbon beer. Knowing that the average consumer often listens to what trendsetters say, marketers go after specific groups of influential consumers.

But they may be missing the mark. Recent research reveals that the most obvious targets for a marketing campaign —loyal customers and tastemakers — may not create the best buzz for the marketing buck. David B. Godes and Dina Mayzlin, assistant professors of marketing at the Harvard Business School and the Yale School of Management, respectively, conducted a field study of a chain store’s word-of-mouth campaign. The store had a small promotion agency recruit 1,073 participants, dubbed “agents,” in 15 U.S. markets. Some agents were solicited from among the store’s frequent-buyer cardholders; others were recruited by the agency but had never heard of the chain before signing up. Agents were asked to spread the word about the chain among friends and acquaintances over a 13-week period and to report each occasion they whispered in someone’s proverbial ear, in exchange for small prizes.

The word-of-mouth campaign definitely worked, say the authors, in that personal communication about the store’s merits, and the corresponding excitement it generated, turned into real sales. But the authors wanted a more precise look at who exactly was doing all the talking. So a key element of the research, as described in a July 2004 Harvard Business School marketing research paper, Firm-Created Word-of-Mouth Communication: A Field-Based Quasi-Experiment, was to identify the characteristics of the most successful agents so that firms can better understand at whom to target their buzz marketing.

Marketers might be tempted, for example, to spur their most loyal customers into evangelizing for them. But the 381 agents who happened to already be loyal customers of this particular chain — the loyalty cardholders — did not create additional sales through the word-of-mouth campaign. Why? Because loyal customers, posit the authors, are devoted fans who have already told their friends and acquaintances about the product. Marketing dollars might be better spent elsewhere. “It’s somewhat counterintuitive on the surface,” says Godes. “But it’s perfectly reasonable when you think about it. Loyal customers are already doing their work for you. So why invest in people who are already doing what you want them to do?”

On the other hand, the word-of-mouth from the 692 nonloyal agents in the study — those who were not card-carrying members of the loyalty program — had a substantial impact on sales. In fact, although the prizes received by agents were valued only at about $15 on average, nonloyal agents generated about $192 in new sales each time they told a friend or acquaintance about the store.

Of course, marketers can’t simply target every consumer who isn’t part of their loyalty programs. They still have to identify the best generators of buzz among nonloyal customers. Conventional wisdom holds that individuals who are considered opinion leaders by their peers — often called evangelists, experts or influentials —are the appropriate people to generate buzz about a product category (food or cinema enthusiasts are two such examples). But this study indicates that opinion leaders, as identified by a survey of the agents, do not generate effective word-of-mouth —unless they are already loyal customers. “Opinion leaders are going to be hard to motivate to talk because they think of themselves as experts, and they’re not excited about talking up products that they don’t love,” says Godes. And if it’s a product they already love, they are already spreading the word, so a campaign directed at getting them to talk might not improve sales.

Though the study casts doubt on targeting loyal customers and opinion leaders, it does suggest a group that word-of-mouth marketers might focus on. The researchers used another survey to determine each agent’s “network density” to measure whether an individual’s propensity to meet new people and connect with friends and acquaintances would be a predictor of word-of-mouth effectiveness. Indeed, the research indicates that the “social butterflies” among nonloyal customers are best at turning word-of-mouth into sales.

The context for the study should be noted when considering these results. The researchers examined a specific marketing campaign that was designed to generate word-of-mouth. As such, the study does not compare the overall effect of word-of-mouth among loyals versus nonloyals. In the end, loyal customers may create more overall word-of-mouth sales, say the authors, even though the marketing campaign had little effect at the margin. Nor does the study suggest that a firm should ignore loyal customers in its overall marketing efforts. Outside of a word-of-mouth marketing campaign, stresses Mayzlin, a firm would still want to take steps to ensure its loyal customers create positive word-of-mouth.

There is more research to be done, the authors are quick to point out. For instance, the results need to be replicated in other product categories with different characteristics, such as those with a higher price. Nevertheless, this study opens a new window of understanding on how a company can use buzz marketing to its advantage.

For more information, contact Godes at dgodes@hbs.edu or Mayzlin at dina.mayzlin@yale.edu.