Viral marketing is thriving: one recent analysis estimates that from 2001 to 2008 firms increased their spending on viral campaigns by 20-fold. Viral campaigns — sometimes called word-of-mouth marketing — are attractive because they allow firms to get closer to consumers than traditional marketing tools by seeding sample products within a target group of potential customers. The payoff comes in increased sales and brand recognition when participants spread the word about a product.
“But there are a lot of things we don’t know about how viral marketing works,” says Professor Olivier Toubia. “Do viral campaigns work better than traditional marketing campaigns? Exactly how does word of mouth spread? How can a campaign be optimized?”
The rise of online social networking platforms may help answer some of these questions. Marketers are increasingly turning to sites that combine the connective features of social networking with product review, often building influencer communities that focus on niche groups, such as parents, women or teens. Firms offer previews of their products to members, and members provide direct, unfiltered feedback to firms. The sites can also help generate promotional buzz while allowing firms to measure interactions and track how recommendations move.
Yet many firms remain unwilling to experiment much. Toubia, Andrew T. Stephen of INSEAD and Aliza Freud ’01 convinced a cosmetics firm that it could gain valuable insights into the mechanisms of viral marketing by turning a launch campaign for a real product into an experiment using both old and new marketing tools.
The firm engaged SheSpeaks, a social media platform, to develop a viral campaign and distribute product samples to 10,000 of its targeted female community members. In addition to receiving a sample, community members also received coupons for discounts off future purchases of the product. At the same time, the firm ran full-page magazine ads and free-standing inserts (FSI) in newspapers, both with coupons.
By tracking codes unique to each tool, the researchers were able to measure and compare redemption rates. The viral sampling branch of the campaign outperformed the traditional campaigns by an impressive margin, returning a redemption rate 12 times higher than that of magazines and about 17 times higher than that of FSI.
Coupon redemption is only a partial measure of a campaign’s success, but Toubia points out that its advantage lies in allowing the researchers and the firm to compare across tools. “It may not be profit or long-term brand image, but we have the benefit of comparing apples to apples,” he says. The viral campaign also allowed the firm to learn more about product interest, effectiveness, whether the participants intended to buy or recommend the product and the impact of influencers’ recommendations on their peers.
One open question has been, who are likely to be the most active members of viral campaigns? Toubia and his coresearchers found that they could predict which individuals would be the most active in the firm’s viral campaign by creating a model from an earlier campaign (for a product from a different firm).
“What really differentiates the active members from the less active members,” Toubia says, “are their social characteristics — and those never depend on the product: how many people they talk to, and how influential their recommendations are with people.”
Once a marketer knows those characteristics, predicting influence in viral campaigns doesn’t really improve with additional information. And, Toubia says, the product matters less than the person in making such predictions. “Before we know the relationship to a product or their feelings about it, we can predict the most active members.”
Given the prominence of social networking and online activity overall, one of the big questions about viral marketing is whether most key interactions — recommending or discussing a product — happen online or offline. Participants in the viral campaign reported far more offline than online interactions. “Even though the online component was central to the campaign,” Toubia says, “online interactions appeared as an extension — not a substitute for — offline social interactions in the campaign.”
While the exact mechanisms might differ, Toubia says, “in today’s world of blogs, e-mail and Facebook, face-to-face interaction still seems to be prominent.”
Olivier Toubia is the David W. Zalaznick Associate Professor of Business in the Marketing Division at Columbia Business School.
Read the Research
"Viral Marketing: A Large-Scale Field Experiment"