What Marketers Misunderstand About Online Reviews

When Jonney Shih told his colleagues that he wanted his contract-manufacturing firm to develop and sell laptops under its own brand name, Asus, most of them thought he was nuts. Asustek had been founded in 1989 in Taiwan, and Shih was now its chairman; it was a successful manufacturer of other companies’ computers and video game consoles. But it had virtually no name recognition among consumers, so how could it compete with players such as Dell and HP? Shih ignored the doubters, and in 2007 an Asus-branded product, the Eee PC, got stellar reviews and became a hit. By 2012 Asus was the world’s fifth-best-selling brand of PC, and by early 2013 its tablets were the third-best-selling brand. Shih’s instinct had been correct: With the growing availability of opinions from experts and users, the importance of a brand name had diminished. Asus is not an anomaly. Companies as varied as HTC (smartphones), Hyundai (automobiles), Euro-Pro (vacuums), and Roku (set-top streaming) have all benefited from a shift in the way many consumers obtain and process product information. In the past, buyers typically made relative comparisons (“Is Brand A better than Brand B?”) or went by the maxim “You get what you pay for.” They were largely dependent on information provided by manufacturers in the form of marketing. Today, thanks primarily to user-generated reviews and people’s tendency to consult social media friends about purchases, buyers have other options. The wealth of peer-to-peer information and the unprecedented availability of expert opinions give them access to what’s known as absolute value—a rich, specific sense of what it’s like to own or use the goods they’re considering. Upending the “Compromise Effect” In 1992 one of us (Itamar) worked with the psychologist Amos Tversky on a study examining how the set of products consumers see influences their purchase decisions. One group of participants chose between two cameras, priced at $169 and $239. Another group was shown a third camera, priced at $469. The $239 camera (now a compromise) was much more popular than the $169 camera among the second group. Including a more expensive option made people willing to spend more—a phenomenon dubbed the compromise effect. In 2012 Itamar and a PhD student, Taly Reich, repeated the experiment, with a twist. They first showed participants other cameras, along with user reviews, on Amazon—and the compromise effect disappeared. Decisions were far more dependent on the chosen camera’s features and on reviews than on its price and features relative to the two other available cameras. This is hard evidence of the changing nature of decision making, which has become subject to outside information and other factors beyond a marketer’s control. Every marketer is aware of the rise of online reviews and other sources of peer-to-peer information, but many neglect this trend and market products much as they did a decade ago. We believe that many companies need to dramatically shift their marketing strategies to account for the rising power exerted on future customers by the opinions of existing customers. We have created two tools to help managers do that: the influence mix and the O continuum. Understand Your Influence Mix Customers’ purchase decisions are typically affected by a combination of three things: Their prior preferences, beliefs, and experiences (which we refer to as P), information from marketers (M), and input from other people and from information services (O). This is the influence mix. Think of it as a zero-sum game: The greater the reliance on one source, the lower the need for the others. If the impact of O on a purchase decision about a food processor goes up, the influence of M or P, or both, goes down. In recent years O has taken on increasing weight in many categories, but plenty of exceptions remain. For example, habitual purchases (such as milk) tend to be dominated by P, while someone shopping for a toothbrush is most likely to be swayed by packaging, brand, pricing, and point-of-purchase messages—all components of M. Companies need to ask: To what extent do consumers depend on O when making decisions about their products? We present the answers as points along the O continuum. The closer your product is to the O-dependent end, the greater the shift in how consumers gather and evaluate information about it. (See the exhibit “How Much Does Opinion Matter?”) How Much Does Opinion Matter? The power of other people’s views varies from one product category to another. The examples below show where several categories fall on the O continuum. Once firms understand where a product falls on the O continuum, they can consider the strategic implications in four realms: Competitive position. In domains where customers depend mainly on O, branding takes on less importance, and newcomers find relatively low barriers to entry as a result—as Jonney Shih’s story shows. This is also apparent in the restaurant business: Research by Michael Luca, of Harvard Business School, suggests that in cities where large numbers of diners rely on Yelp reviews, independent restaurants tend to benefit, while chains and franchises often see their revenues decline. Companies in O-dependent markets can also diversify more easily than others, because new peer-to-peer information can overcome long-held conceptions about what a company is (and isn’t) good at. LG and Samsung have taken full advantage of this capacity, moving beyond their original core products (electronics) into a broad array of tech goods and home appliances. In general, we see greater market-share volatility in domains where customers depend mainly on O. (Witness the swift declines of Nokia and BlackBerry.) Conversely, brand equity and loyalty can protect established players in O-independent domains; brands such as Clorox and Bud Light, influenced primarily by P and M, enjoy relative stability. O is also not of great concern to the likes of Grey Goose vodka and Hermès—brands for which prestige and emotional ties play an important role and quality is a given. Communication. Let’s consider what happens in this arena for products suited to O-dependent purchase decisions. In recent years many camera buyers have turned to ratings and user reviews as their main source of information. This means that celebrity endorsements are less effective than they once were. Banner ads intended to create or reinforce brand awareness are not very successful either, because when it comes time to buy, the weight of trusted reviews usually overrides any residual effect of fleeting exposure to an ad. Instead, companies such as Nikon and Canon should focus on generating user interest in particular products and promoting an ongoing flow of authentic (and positive) content from O on internet retail sites. Consumers are less likely to consult O about purchases that are not very important to them—most people don’t go on Facebook or Twitter to ask “What kind of paper towels should I buy?” or “What brand of detergent do you like best?” So marketers trying to reach O-independent consumers can be guided by some of the old rules, including many traditional M activities. P&G, for instance, can still benefit from persuasive advertising and eye-catching store displays for Bounty and Tide. Market research. Companies in domains that are not susceptible to O can continue to draw on conventional market-research approaches, but those in O-dependent areas need to think differently. Market research usually aims to measure P—it tries to predict the kinds of products consumers will like by assessing their preferences in the past. But as purchase decisions become more reliant on O, rooting around in consumers’ psyches to understand P yields lower returns. For example, a market research study conducted in early 2007—before the release of the first iPhone—concluded that U.S. consumers would not be interested in a “convergent” device that combined the functionality of a cell phone, an MP3 player, and a camera. (Whoops.) What went wrong? The study had measured P, but as soon as the iPhone hit the market and early adopters began gushing over it, people became influenced by O. Instead of measuring individual consumers’ preferences, satisfaction, and loyalty, marketers should redirect resources to the systematic tracking, coding, and quantifying of information from review sites, user forums, and other social media. Product segmentation. A product’s location on the O continuum often varies across customer segments and from country to country. One group of consumers might rely primarily on O, while another might be more attuned to M. And some distribution channels are less conducive than others to O. (Shoppers in brick-and-mortar stores are often more susceptible to M than online shoppers are, for instance.) Companies should analyze different consumer segments and tailor their marketing strategies accordingly. When communicating with segments that rely on M, a company can use advertising to build top-of-mind awareness, persuade customers, and position its offerings—but those strategies probably won’t work for segments that rely on O. Marketers should also bear in mind that the degree to which a particular customer relies on O might vary with situational factors. For example, some of the people who take full advantage of O while shopping for electronics online may come under M’s influence on Black Friday, when ads touting deep one-day-only discounts abound. With not much time to decide or to consult reviews, they may pick up products impulsively, in the belief that “if it’s on sale on Black Friday, it must be a good deal.” No Going Back When we talk with companies about shifting their marketing mix in recognition of the rising power of O, we hear one consistent objection: Growing suspicion (and in some cases, proof) that online reviews are subject to manipulation and fraud. Some marketers believe that consumer reliance on O will decline as more shoppers become wary of deceptive reviews. We disagree. Yelp, TripAdvisor, and others are becoming increasingly sophisticated at weeding out fake reviews, and consumers are developing a better sense of which sites (and which individual reviewers) they can trust. Moreover, we don’t believe that consumers used to the richness of online reviews will ever return to relying on traditional M. Consider two data points. First, 30% of U.S. consumers say they begin their online purchase research by going to Amazon for product information and reviews; second, studies commissioned by Google have found that shoppers consult 10.4 sources of information, on average, before making a purchase. Voracious information-seeking has become deeply ingrained in many consumers, and we can envision no scenario in which they will see traditional marketing as a better provider of product information. The sources of O change rapidly. New review sites and game-changing technologies can suddenly appear. For instance, consumers who use smartphone apps such as ShopSavvy to compare prices can minimize the weight of M on their decisions even on Black Friday. The idea that a new website or app can undercut years of careful messaging may be deeply frustrating to marketers—but it is a reality they must face. As the influence mix evolves, success will come to companies that can closely track the sources of information their customers turn to and find the combination of marketing channels and tools best suited to the ways those consumers decide.

What Marketers Misunderstand About Online Reviews

Partner Center