The Limits of Scale
The value of many products and services rises or falls with the number of customers using them. The fewer fax machines in use, for instance, the less important it is to have one. In industries as varied as credit cards, fashion, and online games, these “network effects” influence consumer decisions and limit the number of companies able to compete.
Strategists have developed some well-known rules for navigating business environments with network effects. “Move first” is one. “Get big fast”—by aggressively growing market share, keeping prices low, and acquiring smaller players—is another. We studied dozens of companies around the world to see whether these rules do in fact lead to success. We found that in many instances the rules did provide a playbook for superior performance. But we also uncovered a disturbingly large number of situations in which the conventional wisdom was dead wrong. And when the rules failed, the reason was always the same: In rapidly scaling, the company neglected to take into account differences among its customers.
In the following pages we’ll explore how both incumbents and new entrants can take advantage of customer differences to challenge seemingly secure competitors and fuel growth in markets with network effects.
Two Dimensions of Difference
According to the textbooks, eBay should own the Chinese market. In 2004 it acquired the largest local online-trading company, EachNet, which enjoyed an 85% market share at the time. EBay’s CEO, Meg Whitman, had witnessed the power of network effects in the company’s U.S. business and was supremely confident: “Ten to 15 years from now,” she said, “I think China can be eBay’s largest market on a global basis as we build up the local trade and the export trade.”
Things turned out rather differently. Taobao, a Chinese upstart owned by the Alibaba Group, completely displaced eBay within a few years. To understand how that happened, let’s step back and think about the people who use such services and products. Our analysis reveals that customer groups can differ in two important ways:
Mutual attraction. Even in markets with network effects, not all customers benefit from the presence of every other customer. Consider two groups of Twitter users, one with an interest in politics and another that is passionate about entertainment. Twitter becomes more useful to the first group as more people share political news and information. Similarly, those interested in celebrities value Twitter more if a larger number of tweets cover entertainment. Some common ground may exist between the two groups—for example, news about the singer Lady Gaga’s political activism would be of interest to both. But generally the political types are strongly attracted to other political types, and the celebrity-obsessed prefer their own kind. The mutual attraction within the groups is very strong; between the two groups, it is far weaker.
When eBay first entered the Chinese market, e-commerce was in its infancy. At the time, technical equipment such as motherboards made up the bulk of online auction purchases, and EachNet, the company that eBay acquired, appealed mostly to technically sophisticated customers. Thanks to strong network effects, eBay’s platform became an increasingly attractive place to buy tech products.
Taobao’s Chinese executives recognized that the company couldn’t compete head-on with eBay in the existing market. So instead they focused on an emerging segment of online auction customers—people on the hunt for clothing and consumer products. Although eBay had a leading position in terms of overall market share, its share of the new segment—which would come to dominate e-commerce in China—was far less imposing. What’s more, eBay’s strong position with techies was no help at all in attracting fashion-focused customers, who were more interested in whether other fashionist as used the site.
Mistakenly assuming that the company had purchased its way to market leadership, eBay’s executives committed a series of strategic errors, ones they might have avoided if they had realized the threat Taobao actually presented. For example, eBay was slow to offer an integrated payment solution, and it insisted on charging customers significant transaction fees. Had its network advantage been real, the model would have made sense—the company with the strongest network effects can typically get away with higher charges (or lower quality). But eBay was not dominant in the emerging consumer market—the mutual attraction between fashionistas and techies was weak—and so its model didn’t fly. In 2006 eBay shut down its business in China.
Asymmetric attraction. In any network, some customers are more equal than others. Shopping malls provide the classic example. Retailers benefit from the presence of anchor tenants, which are powerful magnets for consumers and help drive traffic in all stores. But anchor tenants benefit less, or not at all, from the presence of smaller stores with lesser brands. The attraction between anchor tenants and small stores runs mostly in one direction; it is asymmetric. Fashion offers another example. Many customers like to wear the kind of clothing that they see on stylish people, creating significant network effects. But celebrities have a disproportionate influence on the fashion decisions of regular customers.
In markets with asymmetric attraction, getting big fast is often a poor strategy. Smart firms instead restrict the size of their network to focus on select customer groups that promise to attract everyone else. Threadless, a Chicago-based clothing company that crowdsources its designs, initially restricted its membership to designers of the respected Dreamless community. From a traditional perspective, this move seems puzzling. Companies like Threadless compete on the quality and variety of products. The greater the number of designers, so the logic goes, the stronger the company’s competitive position. But in the fashion market, asymmetric attraction is important: Limiting contributors to just a few respected designers made others hungry to join. Threadless did open up eventually—anyone can submit designs now—but only after the company had established its position as market leader.
Getting users to join a new online community is not easy. Such efforts very often fail, usually because the companies pay little attention to differences among their customers and variations in the strength of mutual and asymmetric attraction among them. Friendster, an early Canadian-based precursor of Facebook, illustrates how a lack of selectivity can weaken a growing network. The company’s service had huge appeal—so much so that Friendster did not have the server capacity (and financial means) to accommodate everyone interested in signing up.
Faced with this constraint, Friendster decided to admit new members on a first-come, first-served basis, adding large numbers of users in the United States and Asia. But because the mutual attraction between these groups was low, this strategy actually made Friendster’s network effects weaker: Asian users did not care much about U.S. users, and vice versa. It would have been better to focus on one region, say customers in New England, and perhaps better still to focus on New England celebrities, in an attempt to exploit asymmetric attraction. The lesson is clear: Indiscriminate growth often undermines the very network effects you’re trying to achieve. That’s why developing a finer understanding of customer differences can lead to better strategic decisions, for both new entrants and incumbents.
Strategies for New Entrants
Dethroning an incumbent is a famously hard challenge for any company, even if the new entrant offers a superior product or service. Exploiting the differences among the incumbent’s customers can help upstarts gain an edge. Let’s look first at a strategy for networks in which the mutual attraction of customers varies.
Pick off the most interested. For much of its history, the New York Times was a regional newspaper with a small readership outside its home area. This was no coincidence. The newspaper business exhibits strong network effects at the local level: More readers attract more advertisers, which attract even more readers. Local interests tend to dominate, and as one or two papers achieve critical mass, they leave the others behind. As a result, most U.S. cities have only a few major papers.
When the New York Times expanded to serve readers nationally, it had to overcome local network effects. It did so by focusing in each new market on the small subset of readers whom it could serve exceptionally well: those with a strong interest in international news. These readers were only weakly attracted to locally focused readers and advertisers. As they switched to the Times, the local competition cut costs by reducing global coverage, prompting additional readers to follow suit. As its overall circulation grew, the paper became more attractive to national advertisers. In some cities, it split the market with the local competition. In others, it put the locals completely out of business.
Create an exclusive experience. EHarmony, an online dating company, used asymmetric attraction to its advantage. Despite Match.com’s leading position in an industry where customers value more choice and tend to flock to the largest sites as a result, eHarmony managed to successfully enter the market by creating an exclusive service for people looking for long-term relationships. In essence, it exploited the fact that for a subset of Match.com’s customer base, some members were of more value than others: Commitment seekers wanted to connect with other commitment seekers; the casual daters in the mix were a nuisance. Because eHarmony screened out the daters, it provided a superior experience for the “committed” types. It entered a smaller market than the one Match.com served but did so successfully, attracting 600,000 users in just five years.
Strategies for Incumbents
At some point dominant players in mature network-based industries face diminishing returns as they attempt to acquire new customers. The results can be extreme: A 2013 working paper by the University of Pennsylvania’s Michael Sinkinson, for example, estimated that the exclusive distribution contract for Apple’s first iPhone was worth more than $20 billion to AT&T. However, the extra customers were worth much less—only around $3 billion—to Verizon, the wireless market leader that already enjoyed a huge network.
When opportunities in existing markets begin to dry up, sustainable growth generally comes from entering adjacent markets, expanding into new geographies, or adding complementary features. Factoring into your analysis the differences among your customer groups can help you make the right choices in each case.
Adjacent markets. These are closely related but not identical markets, and there are usually several to choose from. In most cases, to make the right choice you must assess the level of mutual attraction between your current customers and those in the adjacent market.
Consider the travel services business TripAdvisor. Thanks to strong network effects—the more customers who write reviews, the more valuable the site becomes—TripAdvisor has captured more than twice the market share of its next-best competitor in the “destinations and accommodations” category. From this position of strength, the company considered expanding into restaurants and flights.
TripAdvisor’s principal attraction is the reviews that customers enjoy reading and writing. Such reviews are equally valuable to people choosing a restaurant; however, reviews of past flights on a given route for a given airline are poor predictors of the quality of future flights. Consequently, other people’s experiences are largely irrelevant. Note that the customers in all three markets (hotels, restaurants, flights) are the same people. But what they value about the TripAdvisor experience in hotels carries over to restaurants—and not to flights. This meant that restaurants provided a far better opportunity for expansion.
New geographies. The mutual attraction between existing and new customers also matters when choosing new geographic markets. Consider Wikipedia. Founded in 2001, the encyclopedia now offers more than 4 million entries. There are strong network effects in this industry: the larger the number of entries, the greater the benefits of using Wikipedia.
When Wikipedia sought to expand its offering to foreign-language encyclopedias, it considered a range of markets. One option was to select a large market in which few people speak English and thus are unable to use the English-language version—Japan, for instance. The problem with that approach is that because only a small number of Japanese speak English and not many Americans speak Japanese, Wikipedia can’t leverage the network effects of its English-language version. It would have no special competitive advantage in launching an online encyclopedia in that market.
But not all geographic opportunities have customers with such low levels of mutual attraction. By targeting countries in which English was commonly spoken as a second language and recruiting bilingual authors, Wikipedia was able to exploit advantages from its existing network. Many early contributors to the English-language version were Dutch and German. Those authors were well positioned to help launch foreign-language editions in their home countries; they were already familiar with Wikipedia and its rules of authorship, and they enjoyed contributing to the crowdsourced effort. The bilingual readers benefited from the U.S. network effects directly, and the monolingual readers were served by the local language entries prepared by the bilingual authors. Wikipedia eventually expanded to 285 languages. (See the exhibit “Wikipedia: The English Advantage.”)
Complements. As incumbents expand into adjacent markets and new geographies, levels of both mutual and asymmetric attraction increase, presenting opportunities for focused entrants. But here, the strategic priority lies not in exploiting differences but in finding ways to bridge them.
One of the best ways to do that is by offering a complement (a product or service that adds value to another product or service: for instance, a razor and razor blade). Consider Facebook. How did it manage to become a dominant global enterprise when its network effects, while enormously strong, were largely U.S.-based? Most Norwegians, for example, don’t care that Facebook is the leading social network in the United States, because they do not have many American friends. So why did Facebook displace domestic competitors such as Blink and Playahead?
The answer was social games like Zynga’s FarmVille. These games, which used Facebook as a platform, allowed users to interact with strangers in an engaging manner, bridging the differences between Norwegians and Americans. Moreover, when Zynga developed its next game, it was able to spread the fixed cost of game development over millions of users, increasing development budgets and (it hoped) the quality of the games. Gaming companies that served smaller markets found it difficult to compete.
This is not just a social media story. Complements play an important role in many industries. Take retail banking. ATMs and branches are strategically important because they create network effects—but these effects are fairly local. A Bostonian appreciates her bank’s having an ATM in Chapel Hill, but the benefit is limited because she is rarely in North Carolina. The missing piece is the complements: banking products and services that customers appreciate irrespective of their location. If a bank developed a reputation for superior financial advice, customers might want the institution’s products and services whether or not it had ATMs in the area. For ATMs, customer preferences are heterogeneous; everybody wants them close to home, which favors local incumbents. In financial planning, however, there is far less heterogeneity, which creates an opening for new entrants.
The strength of network effects can be as varied as the human beings who make up the network. A strategy that fails to take into account this variation—ignoring differences in mutual and asymmetric attraction among customers—is likely to fail. It is the rare and lucky firm that prospers simply by being first and getting big fast. In nearly any market with network effects, customers will find the presence of too many of the “wrong” sort of people a distinct turnoff. Welcome to the networked economy—perhaps not so different from the old economy after all.View at the original source
No comments:
Post a Comment