Trillions of dollars of consumer spending have, historically, depended on a few steps. A shopper learns about a product, considers whether to buy it, decides to do so, goes to a shop. If he likes it, he may buy it again. Marketers have long obsessed over each step, and consultants have written treatises on how to nudge people along. E-commerce is already changing the process, but now retailing gurus are imagining a future in which shopping becomes fully automatic.
The idea is that a combination of smart gadgets and predictive data analytics could decide exactly what goods are delivered when, to which household. The most advanced version might resemble Spotify, a music-streaming service, but for stuff. This future is inching closer, thanks to initiatives from Amazon, lots of startup firms and also from big consumer companies such as Procter & Gamble (P&G).
Buying experiments so far fall into two categories. The first is exploratory. A service helps a shopper try new things, choosing products on his or her behalf. Birchbox, founded in 2010, sends beauty samples to subscribers for $10 each month. Imitators have proliferated, offering everything from dog toys to trainers. MySubscriptionAddiction.com, which reviews these services in English-speaking countries, counts 998 new subscription boxes so far this year, up from 284 new ones in 2013. Retailers such as Walmart have followed suit with their own boxes. The scope for such services, however, may be limited. One third of those surveyed by MySubscriptionAddiction.com said they cancelled at least as many subscriptions as they added this year. Consumers, naturally, will delegate purchases to a third party only when they receive products they like. In future, firms that comb purchase histories and search data may be able to send more reliably pleasing product assortments. For now, a consumer who becomes an unwitting owner of toeless socks, which were included recently in a box called FabFitFun, may decline further offers.
The second category of automated consumption is more functional. A service automates the purchase of an item that is bought frequently. Nine years ago Amazon introduced a “Subscribe & Save” feature, offering consumers a discount for agreeing to buy certain goods regularly, such as Pampers nappies. Dollar Shave Club, a male-grooming-products firm, sells razors to subscribers directly, and P&G now has its own, similar service. It is also testing one for laundry detergent.
Amazon is going further. Last year it began selling so-called Dash buttons, designed to be placed around the house to order everyday products—one for Campbell’s soup, for instance, and another for Whiskas cat food (pictured). Investors see this as the first step in its bid fully to automate buying of daily necessities. Already, some manufacturers have integrated Amazon into their devices; General Electric, for example, offers washing machines that shop for their own detergent.
Such services have obvious appeal for Amazon and for big consumer brands. If a shopper automates the delivery of a particular item, the theory is that he is likely to be more loyal. For some brands, the buttons are working especially well: more than half of all the many Amazon orders for Maxwell House coffee in America, for example, are made through the Dash button. Amazon says that across America, an order from a Dash button is being placed more than twice each minute.
But neither Amazon nor the big product brands should celebrate a new era of shopping just yet. Amazon does not release comprehensive data on its automated services, but Slice Intelligence, a data firm in California, reported in March that fewer than half of those with Dash buttons had ever pressed them. One problem may be the e-commerce giant’s prices, which fluctuate often. Another report, by Salmon, a digital agency inside WPP, an advertising group, found that far more British consumers would prefer a smart device that ordered the cheapest item in a category to one that summoned up the same brand each time. That suggests that automated shopping, as it expands, might make life harder for big brands, not prop them up.