Jeff Bezos has been chasing online grocery since 2007.
That August, AmazonFresh debuted on Mercer Island, a suburb of Seattle. For the next six years Amazon tweaked the model. It tried a free loyalty program called “Big Radish” and tested free or discounted delivery based on a customer’s spending. In 2013, AmazonFresh expanded to Los Angeles and San Francisco.
But a decade later, AmazonFresh has yet to become a household name like its sister program, Amazon Prime. Now, Bezos is buying Whole Foods for $13.7 billion.
Online grocery is the holy grail of retail. Groceries are one of the few things that most people buy routinely, and companies from Amazon to Walmart and Target believe that a reliable grocery offering is the key to attracting more customers, and turning their existing ones into better, more loyal shoppers.
Webvan, one of the most spectacular flameouts of the first dot-com bubble, made a bad name for online grocery. The company launched in June 1999 and went public later that year, soaring to a $7.9 billion valuation on its first day of trading. Then, in 2001, it collapsed. Two thousand people lost their jobs. Webvan, it turned out, had never managed to make online grocery sales profitable.
Nearly two decades later, the next round of competitors insist that, this time, things are different. They include AmazonFresh, Silicon Valley startup Instacart, and Walmart, which last year partnered with Uber and Lyft to test same-day grocery delivery. But no one—especially Amazon’s CEO—has yet called online grocery easy.
“[W]hoever your anonymous sources are on this story… they’ve mixed up their meds!” Bezos tweeted earlier this year, responding to a story in the New York Post that claimed Amazon Go, an automated grocery-slash-convenience store the company is testing, would need just three human workers and would yield operating profits margins of more than 20%. “And on the NYPost article, if anybody knows how to get 20% margins in groceries, call me! :)” Bezos added.
Selling groceries is a notoriously tricky business, both on- and offline. For every $100 shoppers spend, stores usually keep only $1 to $3. Groceries are also perishable, meaning vendors always carry the risk that some of their merchandise will go bad before it can be sold. Startups like Instacart and Shipt have tried to get around this—and avoid becoming the next Webvan—by facilitating only the delivery of groceries from existing stores, rather than holding any inventory themselves. “We’re not building all these capital-intensive assets,” Shipt CEO Bill Smith recently told Quartz. This model, of course, sets these companies up as middlemen, making them an additive cost. Consumers are extremely price sensitive when it comes to groceries.
AmazonFresh has so far opted to build its own warehouses, but even for the world’s foremost logistics expert, that approach has proved challenging. Online retail prizes consistency, but groceries—especially fresh ones—rarely allow for that. At its Seattle fulfillment center, for example, Amazon used to waste nearly a third of the bananas it purchased because Fresh only sold the fruit in bunches of five. Workers would discard any bunches of three or four, and tear off and throw out the extra banana on a bunch of six, a research paper by a student at MIT explained in 2015.
Aside from his tweets, Bezos has been relatively tight-lipped on grocery. The first Amazon Go was supposed to open in Seattle earlier this year but has been delayed. Fresh has continued to expand to new markets in the US and recently launched in Tokyo, but Amazon executives have given little insight into its actual performance. “Certainly a business where we continue to work on costs and profitability, but we are finding it’s a very attractive service to our customers, which is what we’re after,” Brian Olsavsky, Amazon’s chief financial officer, said on the company’s third-quarter earnings call last year.
But on one point there should be no doubt: Bezos has his sights set on online grocery, and he’s determined to win. It may happen with AmazonFresh, AmazonGo, Amazon Prime Now, or something that Amazon has yet to try—like an acquisition of Whole Foods, whose hundreds of US stores are ready-made distribution centers. Online grocery sales totaled $20.5 billion in the US last year and analysts believe it could be a $100 billion market by 2025. If buying Whole Foods is what it takes to make Amazon a household name in that market, $13.7 billion will have been a bargain.