Menu Close

Amazon’s bottomless appetite

file zvylji.

Amazon is shrinking time and eating up the planet. Once a small online book retailer, in 2018 it took over Whole Foods Market, a leading US organic grocery chain. That’s almost 500 stores in one mouthful. In Paris, the firm is vacuuming up distribution of goods sold by the Monoprix supermarket chain. There is no longer a need to queue at the delivery counter or browse, then stay at home waiting for your hummus or dish soap to turn up.

Nowadays everyone has heard of Amazon’s website and its express delivery service, but we pay less attention to its appetite for competitors and former associates.

Let’s start with a reminder of Amazon’s vital statistics. It weighs in at nearly $200 billion in annual revenue, with a $760 billion market valuation, the highest in the world, ahead of Google and second only to Apple. Globally, it occupies 140 square kilometres of warehouses and logistics bases, equivalent to the surface area of Cambridge and Oxford, England, combined. Every day it dispatches some 1.6 million parcels, with a workforce of more than 500,000 (25% of that of its competitor Walmart).

Wreaking revenge with diapers

Now for a tale of diapers. Once upon a time there was a start-up called Quidsi, which achieved spectacularly rapid growth selling baby products online. Young families on the US East Coast flocked to its site,, in search of wipes, diapers, milk powder and such, all delivered for free. Amazon expressed an interest in buying Quidsi, but the start-up did not relish the idea of being swallowed in one gulp.

Jeff Bezos, founder of Amazon. Doc Searls/Flickr, CC BY

Shortly afterwards, or so the story goes, the Seattle giant lost its temper. It offered a 30% discount on diapers for customers who agreed to place a monthly order. It rolled out the “Amazon Mom” programme, which included a year’s free delivery of baby products within two days of placing an order. Quidsi alleges that in so doing, Amazon accepted quarterly losses of up to $100 million on diapers alone.

Time passed and Quidsi’s growth curve flattened out, while those of its increasingly large rival soared. The upstart fell out of favour with investors and when Amazon renewed its interest, Quidsi sold out for $545 million. A few years later Amazon shut down and Quidsi’s other sites.

A serious affair. BoogaFrito/VisualHunt, CC BY

Standing up to Amazon is expensive

The moral of this tale of diapers is pretty obvious: it’s hard work competing with a giant like Amazon. Some analysts claim that the Seattle-based company was openly predatory, a form of behaviour normally prosecuted under antitrust law. Once it has eliminated the competition with rock-bottom prices and minimal or even negative profit margins, it can raise consumer prices again and offer less attractive terms of sale.

Others hold that Amazon’s strategy simply borders on intimidation and its sales ploys are perhaps a bit aggressive. Furthermore there is no clear proof of any deterioration in consumer welfare. After close scrutiny, the Federal Trade Commission (FTC) approved Amazon’s takeover of Quidsi, asserting that it would neither hurt competition nor jeopardise consumer interests. Lacking a sufficiently detailed grasp of this case, we can only accept their verdict.

Being a listed Amazon seller can be expensive too

We should add that it is not always a good idea to be listed as an Amazon seller-partner, here again due to its size. But first a word of caution: if you are not a retailing specialist, the meaning of various terms demands clarification. For this purpose we shall stick with baby products. Amazon sells its own brand of Mama Bear diapers, as well as reselling Pampers and Lillydoo goods, among others. Finally it showcases the products of other specialist manufacturers and traders.

Brick-and-mortar or online retailing generally uses one of three models:

  • The first is based on partial vertical integration: the retailer defines a product’s characteristics and owns the brand under which it is marketed, for example store brands used by mass retailers such as Tesco (UK) or Carrefour (France).

  • In the second and most frequent model, retailers resell products purchased from third parties, such as Procter & Gamble’s Pampers diapers.

  • In the third case the retailer is a marketplace, a “pure” intermediary between buyers and sellers thanks to its website or store: think of the mini-stores for Apple phones or luxury cosmetics in big shops.

Amazon uses all three. In particular it is both a reseller and a marketplace. Originally it only resold books, subsequently adding a few other products. But commissions paid by third parties now account for 20% of overall revenue. At the same time its business as a marketplace adds up to half of all transactions. In other words Amazon does not itself purchase 50% of the products sold on its website.

In an Amazon warehouse (2014). Forgemind ArchiMedia/Flickr, CC BY

This is an essential point for two reasons. First, a marketplace is characterized by very specific network effects: the more seller-partners there are at a given site, the more attractive it is for consumers to shop there; similarly the more visitors a site attracts, the greater the incentive for sellers to be represented there. The same, well known mechanism is at work in a platform economy, with its markets controlled by one or just a few firms. A reseller is neither a marketplace nor, as economists put it, a two-sided market.

Second, the risks involved are not the same. Resellers purchase the products they sell. It’s just too bad if they fail to find a buyer. However they can drop their price or have a clearance to unload excess inventory, unlike a platform which only acts as an intermediary. In short there is a big difference between a contract with a supplier to purchase goods for resale and a partnership agreement covering sales on a marketplace. The economic theory of contracts demonstrates the extent to which these two options differ in terms of risks, incentives and investments.

A shifting strategy

So it is a strategic decision when a firm positions itself as a reseller or a marketplace. Depending on which products it is selling Amazon has opted to adopt one or other stance. Powerful models explain why it makes these decisions. It tends to prefer to act as a reseller for popular products. For example Amazon generally purchases and resells successful DVDs, whereas third-party partners are left to handle less popular items.

What is even more interesting is that Amazon shifts its position over time. The Seattle-based firm may opt to sell such and such a product directly, whereas in the past it was only available on the marketplace from third parties.

A fascinating econometric study focuses on how Amazon poaches on its partners’ product spaces. Over a 10-month period the authors of the study observed that a shift of this sort affected 3% of products sold by third parties. Nor was this a haphazard occurrence. Amazon focused on goods with a higher price tag, stronger demand and more favourable customer ratings. In other words Amazon moves into markets with a good expectation of higher added value.

The packages arrive. Public.Resource.Org/Flickr, CC BY

Two other results confirm this conclusion. First, Amazon tends to resell products with low delivery costs, which is hardly surprising. The shipping terms and conditions it offers when operating as a reseller are most attractive to customers, but very costly for the firm (see below). So it stands little chance of making a profit on a product the price of which barely exceeds the cost of shipping.

Second, Amazon is less likely to start reselling products that it already stores, packs and delivers. For these products the cost-benefit balance for Amazon is not the same. If it switches to reselling such goods it loses the revenue from inventory management services, paid by the relevant third party, as well as the commission for using the marketplace.

Prioritising growth over profit

Last, the study shows that when Amazon enters the product-space of third-party resellers it often discourages them from continued trading. They are on average more likely than other resellers, not in direct competition with Amazon, to stop selling altogether or reduce the range of products sold on the marketplace. On the other hand when Amazon starts reselling a product it does not increase the retail price. In fact demand increases, because customers pay less for shipping than before. In other words, true to its determination to prioritize growth over profit (see the figure below). Amazon shows no inclination to take advantage of the situation and upset its customers.

It goes without saying that Amazon’s selective approach to the products it resells is based on a mass of data and highly sophisticated processing. Amazon finds out almost all there is to know about third-party sales and products: price, volume, destination, delivery terms and cost, customer ratings and so on. As a former executive explained:

“If you don’t know anything about the business, launch it through the Marketplace, bring retailers in, watch what they do and what they sell, understand it, and then get into it.”

Cambridge University Press has published a book by François Lévêque, “Competition’s New Clothes: 20 Short Cases on Rivalry between Firms”.

This article was translated from the original French.

Want to write?

Write an article and join a growing community of more than 186,800 academics and researchers from 4,994 institutions.

Register now