Google, Amazon and the Myth of Network Effects

Tim Wu, who coined the term Network Neutrality, writes

The Internet has long been held up as a model for what the free market is supposed to look like — competition in its purest form. So why does it look increasingly like a Monopoly board? Google “owns” search; Facebook, social networking; eBay rules auctions; Amazon, retail; and so on.
Internet industries develop pretty much like any other industry that depends on a network: A single firm can dominate the market if the product becomes more valuable to each user as the number of users rises. Such networks have a natural tendency to grow, and that growth leads to dominance.

Fred Wilson, of Union Square Ventures, argues

Internet is a network and the dominant platforms enjoy network effects that, over time, lead to dominant monopolies. We see that with Google today. Google’s global search market share is around 70%.

An article titled, The next Internet monopoly: Uber, the transportation network, states

By design, the Internet trends towards monopoly: There’s one major social network (Facebook), search engine (Google), encyclopedia (Wikipedia) business network (LinkedIn), classified-ads website (Craigslist), retail site (Amazon), auction site (eBay), and micro-blogging service (Twitter).

Monopoly is a recurring theme for Internet businesses. As Fred Wilson and Tim Wu argue, this is attributed to network effects. Wikipedia (a monopoly in itself) defines Network effects as the effect (or benefit) that one user of a good or service has on the value of that product to other people. This effect can be understood easily in the context of a Facebook. Having your friends on Facebook, increases the value of the product for you and leads to more people joining in. A competitive product which does not have your friends is surely inferior and loses users. The network effect can lead to everyone using the same product creating a monopoly. Similar examples can be seen in the context of LinkedIn, WhatsApp, Twitter etc.

A variant of network effects are two-sided markets. Two-sided markets, are economic platforms having two distinct user groups that provide each other with network benefits. Developers choose to develop applications for an operating system with many users, while users choose to adopt an operating system with many applications. This has been observed for Windows, iOS and Android. Each side that chooses Android benefits the other user (as developers develop more applications) creating a network effect for Android.

The four largest tech businesses by market cap are Google, Microsoft, Facebook and Amazon. Network effects significantly contributed to the rise and dominance of Microsoft and Facebook.

PageRank, the research thesis of Larry Page and Sergey Brin, was the idea that counting the number and quality of links to a page was a good estimate of its importance. Based on PageRank, Google developed a search engine. Two years later, in 1999, it tried to sell itself to Excite for $1 million. Excite rejected even the negotiated down offer of $750,000. A December 2000 BusinessWorld article says,

Google has captured 25% of the search-engine market, says consultancy Gartner Group. So, where’s the business model? It boasts 100 co-brand partners, such as The Washington Post and Netscape, that have selected Google as an embedded Internet search engine on their site. Most of these co-brand partners pay the company $600 to $2,000 per month in licensing fees.The company has also instituted a pay-for-play scheme called Adwords that allows an advertiser to purchase a word and place a small text ad on the page whenever that word is mentioned in a query.
An even bigger stamp of approval came in June, 2000, when portal Yahoo! replaced Inktomi’s search engine with Google. “If they can hold on to the Yahoo business, it may prove as powerful for them as it proved for Inktomi initially,” says Gartner senior analyst Whit Andrews.

In Google, a user did not benefit from more users. There was no network effect. Google Search has been a monopoly in most markets. If not network effects, what led to the monopoly.

Search was not a cherished business in 2000. Though there was usage, there were many providers and the business model was restricted to being a product licensors. However, in the succeeding years, two things happened. Search became the start point of the internet and Google developed a killer business model. Also, most search providers served all the users irrespective of geography, demographics etc. So, when Google with its relentless focus on search, cracked open the model it captured all the users. Exceptions were places like China and Russia where policy and language resulted in Baidu and Yandex.

In 1995, Jeff Bezos, set about creating the world’s largest bookstore. Amazon, within the first two months of business, sold to all 50 states in the USA and over 45 countries. And the “Earth’s Biggest Bookstore” carried almost every book in print in its catalogue. Amazon launched Auctions in March 1999 and a fixed-price marketplace business, zShops, in September 1999. Auctions and zShops evolved into Amazon Marketplace, a service launched in November 2000. Talking about Amazon Marketplace, Jeff Bezos in his annual letter for 2014 writes,

Marketplace’s early days were not easy. First, we launched Amazon Auctions. I think seven people came, if you count my parents and siblings. Auctions transformed into zShops, which was basically a fixed price version
of Auctions. Again, no customers. But then we morphed zShops into Marketplace.

Network effects are not associated with a bookseller like Barnes and Noble. Amazon user’s did not benefit from other users. There were scale benefits of having large number of users, similar to Barnes and Noble. The two-sided markets of customers and suppliers launched five years after the launch of Amazon. Like in the case of Google, network effects does not seem to explain the monopoly attained by Amazon.

In the heydays of the internet boom, lots of ecommerce launched — Pets.com, Webvan, eToys. With the bust of 2001, ecommerce as a new industry lost sheen and except for Amazon none of the others survived. Amazon has had to fight concerns about the viability of ecommerce. Also after the Nasdaq crash, Amazon has not faced any competitor. Zappos and Diapers.com who made some headway were acquired by Amazon.

Two factors seem to be at play here :

  1. Imagine a company that comes up with a completely new product. A product that did not exist before. Product Market fit is achieved. Usage grows. As a new product, business model is uncertain. Investors struggle to define market size. Others wait for it to fail. Each milestone is viewed with bewilderment. No one is sure. It seems foolhardy to challenge it. Nobody allocates large capital against it. Soon it attains escape velocity. The company has become the industry. Imagine, if everyone knew that search and ecommerce would be new big industries. Imagine they also knew that they would be hugely valuable. Would the others have responded the same way. Would Google and Amazon have become monopolies. What Google and Amazon have demonstrated is the moat of innovation. A new industry creating product is left unchallenged. Nobody knows how big and valuable it can become. At each step it seems near the peak which discourages competition.
  2. Internet companies cover entire geographies and customer segments. This ensures that there are no easy holes for others to fill. The ubiquitous approach enabled by the Internet and religiously adopted by Google and Amazon widened the Moat of Innovation creating a monopoly.

The 1994 IPO of Netscape heralded the internet boom. Twenty years later, internet is a well established industry. Consumer adoption and behaviours are well understood. Business models have also been proven. New path breaking innovations will continue to happen. However, at a base level, internet is taking the contours of an industry. The expansion of the existing internet products will be driven more classical tools like capital and business models than innovation. The rise of Uber clones — Didi Dauche in China, Ola in India, GrabTaxi in South East Asia and EasyTaxi also indicate the readiness to embrace any new product quickly.

The myth of network effects gave the businesses a halo that could not be challenged. It is always more difficult to replace the leader. As the myth is busted, for businesses that are not built around network effects, its likely that competition will start chipping away through geographical differentiation or product segmentation.