P&G’s Tremor division claims to have cracked the code on word-of-mouth marketing using a concoction of old metrics, new approaches and a heavy dose of secret sauce
By Samar Farah
Pharmacists don’t dream of stumbling on immortality potions, chemists don’t waste hours trying to turn lead into gold, and NASA physicists don’t tinker with time machines.
But a team at Procter & Gamble has spent the past four years trying to hatch the closest thing to marketing magic: natural word of mouth, or what happens when one consumer talks to another consumer. The result is marketing arm Tremor, which claims to have cracked the code on word-of-mouth methodology. Tremor maintains a network of 250,000 carefully screened American teens who volunteer to be the first among their peers to gaze at new products and ideas, with an unspoken expectation that they will share the news with their circle of friends. The success of its teen-based efforts has led Tremor to begin exploring another target audience for its budding word-of-mouth business: moms.
The element of fantasy behind such an endeavor apparently is not lost on Tremor’s creators. Consider the name. A tremor can be an incredibly elusive blip of emotion or thought that evaporates. On the Tremor website, the homepage pulses with the promise of insider thrills.
But despite the mysterious overtones, P&G holds that the Tremor formula is straightforward marketing science: a combination of old metrics applied in new ways, technology-enhanced research and innovative approaches to building consumer relationships. Which is not to say that P&G is willing to completely lift the veil on its bag of tricks. But in an interview with CMO, Tremor CEO Steve Knox shared many of the P&G research insights and the resulting techniques. He also talked about the knowledge gaps in word-of-mouth marketing and the direction he believes it will take in the future.
“The early years at Tremor were spent trying really hard to understand how word of mouth actually works in the marketplace,” says Knox. The work is not over, he adds. “Word of mouth is not a static science; it’s a dynamic science.”
Some skeptics say word of mouth can’t be a science at all—that merely attempting to pair word of mouth and marketing is oxymoronic. Other detractors say that bringing word of mouth under a microscope and trying to reproduce spontaneous puffs of brand advocacy will eventually lead to a public that’s even more mistrusting and cynical than the one currently challenging marketers. But when P&G, the world’s largest advertiser, creates not just a line item but an entire division devoted to generating word of mouth, it can drown out the skeptics.
Tremor is fast gaining clients. Having started in 2001 with campaigns for its parent company’s products as well as external brands, Tremor now runs word-of-mouth programs for movie studios, including DreamWorks, and brands such as Coca-Cola and Toyota. Knox, who politely declines to name (or confirm) any of Tremor’s clients, says that 80 percent of the group’s campaigns are for non-P&G brands.
P&G is not the only company trying to turn word of mouth from a desired effect into a reliable distribution channel. The Word of Mouth Marketing Association (WOMMA), formed last year, has grown to the tune of 150 members and sold out its first conference in Chicago in March. As Tremor and similar models evolve, most major companies are, at minimum, paying close attention to the trend.
“We’re not nearly as far down the path as P&G,” says David Dickey, director of consumer e-business at Sprint. “We haven’t made a determination of where word-of-mouth marketing is going yet. But we want to make sure we have a very good understanding of the space, so if it does continue to blossom and move forward, we are ready to take advantage.”
Fueling the Tremor engine are the 250,000 teens that P&G calls “connectors.” One major insight from P&G’s initial research, Knox says, is that Tremor’s connectors “exist throughout the product adoption curve.” That sets them apart from trendsetters or early adopters, those consumer warhorses of Tipping Point fame. Although trendsetters and early adopters are quick to glom on to new ideas and products, they are not necessarily avenues for successful word of mouth; in fact, some trendsetters might be cul-de-sacs of buzz, hoarding secrets that distinguish them from peers.
A connector, by contrast, is anyone—even the last person to find out about something—who always taps the nearest shoulder to point out a new purchase or a cool song or TV show or movie. They are people with “really broad and deep social networks and a deep propensity to want to talk about ideas,” Knox says.
Because the success of a Tremor campaign ultimately hangs on the lips (or the keystrokes) of these minions, P&G spends a lot of time developing ways to identify them. The first step is to draw teens to Tremor.com. Once on the site, visitors are prompted with questions (such as, “How many people do you talk to on a daily basis?” and “How do you feel about buying new products?”) that screen for eight character traits. Knox shares what he calls the three primary characteristics: inquisitiveness, connectedness and persuasiveness. A typical connector, for example, will have 150 to 200 names on her instant messaging buddy list.
At this stage, the door swings open for roughly 15 percent of applicants, who are greeted as new Tremor members. The rest get a polite, “Thank you for your interest in Tremor. Unfortunately, we have enough applicants at this time.” The chosen ones are then put through what Knox refers to, with an insider’s apology, as “boot camp.” Tremor baits the connector candidates with a series of ideas and opportunities: “Think like a Hollywood Heavyweight,” exhorts the subject line of one Tremor e-mail, for instance, which asks for teen input on a new idea for a movie.
Meanwhile, P&G staff watch anonymously, behind the curtain of the Internet, to see whether the teens’ actual behavior matches their claimed behavior. In short, what do they do with their first exclusive morsel? What follows remains part of Tremor’s secret sauce. Between 8 percent and 10 percent of the original applicants will attain connector status (or, typically, 1 percent of a target audience.) After working their way into the inner sanctum, these official Tremor members are unaware that they are still under evaluation. It’s a detail that Knox admits to readily but which casts a small shadow on Tremor’s claims of transparency—perhaps the main point of contention in any word-of-mouth campaign.
Knox’s conviction regarding the influence of this elite group suggests that connectors are a force ignored at marketers’ peril. Indeed, brands like Friendster and word-of-mouth agencies such as BuzzMetrics are elbow-deep in tracking connectorlike consumers—alternately referred to as influencers, transmitters or bees. However, other marketers looking to practice word of mouth are concerned that these influencers are being assigned too much, well, influence.
The debate usually pits Tremor against BzzAgent, an agency whose client list includes Kellogg, Ralph Lauren and Anheuser-Busch. BzzAgent has no highly evolved, proprietary system for accepting volunteers—it’s come one, come all (so far, more than 86,600 have come). BzzAgent recruits raise their hands virtually, signing up online to test new products and deliver diligent diaries of their conversations with friends, family and others about those products.
What separates BzzAgent from Tremor is the belief that any consumer—from a homemaker to a CEO—potentially has the itch to share a bit of edgy market gossip, and that most people happily collect information from a variety of folk. “Influencers are no more effective at generating word of mouth than everyday people,” says Dave Balter, CEO of BzzAgent. “Our agents are everybody.”
Wherever one may fall on the influencer question, most marketers agree that word of mouth is useless if the person delivering the message is not sincere. For Tremor, its carefully screened connectors are just half the equation for credible word of mouth. The company has also devised methods to groom a product for an optimal word-of-mouth campaign.
Connector relationships are forged online, but in most cases, the bond between a connector and a product takes place offline, with a snail-mailed goody kit that might include stickers, DVDs or product samples. These packages, and the entire campaign, are vetted in meetings between Tremor and brand representatives for two key elements: “advocacy” and “amplification.”
The advocacy and amplification model was constructed from a mix of academic theory, psychological insights, existing P&G data and Tremor’s proprietary research. In Tremor lingo, advocacy is what happens when a connector naturally experiences a product and likes it enough to talk about it with her peers. “When connectors are exposed to a new idea,” explains Knox, “the first question they ask themselves is, ‘Is this idea worth my advocacy?’ It’s their social currency on the line, so it has to be a product that they at least believe in.”
Tremor doesn’t let the fickle emotions of high school hallways determine whether a connector responds to a product. “We have a way of finding what the critical advocacy component is of a brand and what would cause a connector to advocate it,” says Knox.
The second element in Tremor’s word-of-mouth model, amplification, means that a product message or experience is easy to talk about, that it lends itself to casual mentioning in varied conversations. “It’s very difficult to find a word-of-mouth concept that has both high advocacy and amplification,” says Knox.
As a result, Tremor often tests several messages for both levels on a group of connectors before a campaign launch, in search of the one that actually works. Explains Knox, “It allows us to go back to a client before we ever go to market and say, ‘See these eight ideas over here? The connectors won’t talk about them. But see this one? They talk about it like crazy.'”
The key to a good Tremor campaign, then, lies in plenty of data points before a campaign even goes out the door. The idea is that a finely regulated petri dish will create lots of word of mouth. BzzAgent’s specialty, by contrast, is a well-oiled feedback system. BzzAgent trains what it calls “communication developers” to sift through reams of agent reports, about 4,000 to 7,000 a week, and to respond to each individually. This correspondence is eventually funneled into graphs and charts for clients.
Prelaunch calculations aside, Knox is keen to stress that the Tremor kits that reach connector doorsteps do not include a scripted dialogue or talking points about the product. In other words, Tremor teens’ interaction with a product is spontaneous and authentic—even if Tremor has done everything possible to increase the probability of a positive reaction. Elements of advocacy and amplification “are just something that causes the teen in the naturally occurring setting to say, ‘Hey, did I tell you about product X?'” says Knox.
His choice of words, “naturally occurring setting,” might strike some marketers as incongruous with a glossary of terms such as “amplification” and “connector.” Is it really possible to break down into discrete parts the human impulse to gab, re-create that impulse and then legitimately call it “natural”?
Self-conscious word of mouth is not really word of mouth, argues Nicco Mele, CEO of EchoDitto, a technology and Internet strategy company that creates online communities and helps raise awareness and funds for its clients. “The most successful campaigns are unaware that they are marketing campaigns,” says Mele, the brains behind Howard Dean’s 2004 presidential campaign website, which was lauded for its successful use of blogs and other Internet tools to build a grassroots following.
Knox is very aware that connectors might grow weary as the novelty withers and routine and predictability set in. “What keeps me awake at night is overdoing the [connector] panel,” he says. “I need to maintain a relationship with these connectors [to encourage them] to stay involved.” To keep them from tiring, Tremor limits connectors to 20 campaigns a year.
Tremor has other ways of warding off connector fatigue. “Influence campaigns” solicit ideas from Tremor teens during a product’s prelaunch phase as well as for existing products. Inside the Tremor website, P&G rallies teens with messages about their imprint on products. “You told Crest what products you want,” congratulates one headline. “You helped Coca-Cola choose a message in a bottle sent to nearly 100 million people!” reports another, after Tremor teens submitted slogans for Vanilla and Cherry Coke under the theme “bold life of the party.”
These types of campaigns, according to Knox, have inherently high levels of advocacy and amplification because the teen has been linked in a personal way with a brand. “You’ve got connectors saying to all of their friends, ‘I helped pick the music for that commercial.’ It causes the product to be talked about in naturally occurring conversations.”
There’s that word again: naturally.
Return on WOM
Tremor and other word-of-mouth agencies are attracting marketers frustrated with returns on print and TV ads and unsure how to reach media-clogged consumers. That’s why Rick Pascocello, vice president of marketing at the publishing firm Penguin Group, took the plunge four years ago and offered a Penguin novel as a pilot case for a BzzAgent campaign. Pascocello says he is now on his 24th word-of-mouth campaign with BzzAgent.
But as marketers rush to this new medium, they’re not necessarily trailing better ways of measuring return. The science behind Tremor word of mouth, for example, pretty much ends where the teen connector’s advocacy stops. “I can measure Tremor to the connector and to a lesser degree from the connector to their first set of friends. What happens after that, we become relatively blind to,” admits Knox.
Others lay claim to advanced word-of-mouth measurement systems. Agencies such as Intelliseek have developed capabilities to track a brand’s word of mouth online, raking through 11 million blogs, message boards and other online communities. Friendster, an online social network that boasts 16 million members, claims it can measure second- and third-degree levels of word of mouth by tracing why and how members contact one another online. BuzzMetrics, which provides a range of word-of-mouth research and planning services, has developed a syndicated service in which subscribers receive quarterly reports and briefings on the Internet activities of flagged influencers in segments such as the nutrition industry.
It has sold the service to seven of the largest food companies.
“The fact that tens of millions of consumers are now conducting WOM activities in a huge, digitally archived environment opens huge possibilities. It allows for the kind of accountability marketers need,” says Jonathan Carson, president and CEO of BuzzMetrics.
Knox points out that most of Tremor’s clients are less concerned with monitoring how often their brand echoes in electronic caves than they are with tracking sales activity in the areas where WOM campaigns are taking place. Consistent spikes in sales, in fact, are what keep Pascocello returning to BzzAgent.
“I don’t look at all of BzzAgent’s charts and graphs as much as a lot of their clients,” he says. Far more interesting to Pascocello than the complexities of the BzzAgent feedback process are the results. Take The Art of Shen Ku, a book Penguin published Sept. 11, 2001. The book gathered dust for a year until Pascocello, having exhausted print and promotional avenues, decided to try a word-of-mouth campaign in hopes of reviving it. Six months later, sales doubled. “If I had a choice to run one print ad or a buzz campaign,” he says, “I’d do a buzz campaign.”
These types of success stories are leading word-of-mouth innovators into new areas. Comfortable with its methods for the teen audience, Tremor is now branching out into the “moms” segment, which links more closely to P&G’s own product lines. Knox says Tremor will need to alter some of the screening questions—for example, the number of organizations a mother belongs to rather than the number of buddies on her IM list—but the basic definition of a connector will hold. Tremor is also testing alternative recruiting methods to capture this less tech-savvy segment that spends more time offline than on.
And Tremor and its clients are casting their eyes to what they believe is the next application of word of mouth—consumer retention. Knox maintains that the advocacy and amplification model can form the foundation for a brand loyalty program galvanized by word of mouth. “What are the triggers that will cause a connector to talk about this product four months from now and nine months from now and 18 months from now?” Knox asks. “We’re working with a couple of clients now on using word of mouth to build long-term loyalty campaigns that extend out over many years.”
And no, he won’t say who those clients are. But if the current model is any indication, they will receive a generous serving of Tremor lingo, influencer creed and buzz forecasts—the stuff of word-of-mouth science.
From CMO magazine (a publication of CXO Media, Inc.), July 2005