Amazon famously began life as an online bookseller, but 20 years later there are few products you can’t buy on the site. For many goods, the company acts as a traditional retailer, earning a profit through markups. Other goods are offered by third-party sellers—merchants who pay to use Amazon as a sales platform, much as people pay to sell items on eBay. Platform business models, which use technology to link buyers and sellers, have never been hotter. Examples include Uber and Airbnb (which link passengers with drivers and travelers with lodgings, respectively) and technology companies such as Facebook, Apple, and Twitter (which open their ecosystems to outside software and app developers). Venture capitalists love the platform model, because it lets companies scale up with limited funds: Airbnb, for instance, was able to grow far more quickly as a matchmaker than if it had had to build and operate hotels.
But when Feng Zhu, an assistant professor at Harvard Business School, began studying the platform model, he heard a repeated worry from Amazon’s third-party sellers: What happens if, instead of just matching buyers and sellers, the platform decides to offer competing products itself? The more Zhu talked with sellers, the more he saw that this wasn’t a hypothetical question. “Their view is that the platform owners essentially run the platform as a lab, letting people innovate and compete against one another, and then cherry-pick the best products for themselves and capture the value,” he says.
In some respects the problem is an old one. For instance, supermarkets and drugstores sell their own private-label products alongside branded versions, and some fast-food companies open stores that compete in locations with franchises. But in an increasingly digital economy, the complex relationship between platform operators and sellers is especially likely to end up in frenemy territory. For example, the Meerkat app, which allows live video streaming on Twitter, was the hit of the 2015 South by Southwest conference, but Twitter soon pulled Meerkat off its platform in order to promote Periscope, a similar app it had recently acquired. Weather apps, password apps, and flashlight apps have all struggled since Apple made their functions standard features of its operating system.
To understand how the dynamic is playing out on Amazon, Zhu and a colleague identified 164,000 products sold exclusively by third parties on the company’s platform during June 2013—that is, items Amazon itself did not sell. Ten months later the researchers examined those product categories and found that Amazon had begun directly selling items in 3% of them, for a total of 5,000 items. Although 3% may not sound like much, Zhu says that it’s a significant amount over such a short interval and that it suggests Amazon may move into competition with many of its third-party sellers over the long term. The analysis also showed that products sold directly by Amazon often became the default option in search results.
The researchers then sought to understand why Amazon had chosen those particular products for direct sales. They determined that the company was more likely to offer products directly when those products had strong demand relative to other items in the same category; when there were numerous third-party sellers (signaling that it was easy to source the goods from manufacturers); when shipping costs were low (especially important because Amazon often offers free shipping); and when prices were relatively high. It was less likely to enter categories in which the third-party sellers used the Fulfillment by Amazon program (whereby sellers pay Amazon to manage inventory, orders, and shipping).
In theory, a company might move into a product’s space because the third-party seller has done a poor job with customer service—a scenario in which the benevolent platform operator is looking out for customers, not just seeking additional profits. But after analyzing customer reviews of the products Amazon had chosen to offer itself, the researchers concluded that that wasn’t its reason; most customers seemed happy with the products they had bought prior to Amazon’s entry. “The dominant pattern here is that Amazon is more likely to target products that are performing well on Amazon.com,” Zhu says. (Amazon declined to comment on Zhu’s research.)
When Zhu has presented his findings at other universities, academics have asked: What’s the impact on consumers? Amazon’s entry into a category generally doesn’t change product prices, Zhu says; but because of the frequency of free shipping on Amazon, total costs to consumers often fall. However, he worries that over time, Amazon’s behavior might erode sellers’ willingness to introduce innovative products on the platform. “We found that small sellers affected by Amazon’s entry are discouraged from growing their businesses on Amazon Marketplace,” he says.
If you’re a third-party seller, Zhu’s findings suggest that some wariness and defensive action are in order. Keep in mind that when selling on Amazon, you’re a middleman—a position often subject to disintermediation. The researchers suggest steps that can reduce the odds of Amazon’s entering your market. Sign exclusive deals with suppliers, or manufacture some products yourself. Focus on a large selection of niche products (Amazon appears to target widely popular items). “From day one, as they build their business model, third-party sellers need to design and position their products to minimize their risks,” Zhu says.
About the Research: “Competing with Complementors: An Empirical Look at Amazon.com,” by Feng Zhu and Qihong Liu