Amazon Edges Closer to Fully Automated Retail

With the Dash button and service, the tech giant wants to make your house do the shopping for you.

Ian Bogost

I press the white, round button on the Gatorade-branded Amazon Dash Button in my palm. It’s lozenge-shaped, about three inches long, faced with a black bezel sporting the sports drink’s logo. The smooth, concave button is a pleasure to push. It scallops just a little, modestly, less than you would have expected. Then a white LED flashes, and you have ordered Gatorade—or Tide, or Depends, or Kraft Mac & Cheese, or a dozen other household products.

“When will it get here?” my daughter asks. The whole family flanks her. We have all gathered to press the button for the first time.

“In two days,” I answer. “Just like any Prime order.” She furrows a brow, nonplussed. Even though the tactical pleasure of pressing the button still lingers, her silent disappointment is contagious. A magic button that delivers Gatorade on-command is less magical when all it does is place an Amazon order on your behalf. Lost magic quickly gives way to farce: What, today’s sluggards can’t even sit upright enough to operate computers for ordering commodities to be delivered to their stoops?

Dash Button has oscillated between magic and farce since it was revealed almost a year ago. Even the announcement was ambiguous: it came just before April Fool’s Day, forcing my colleague Adrienne LaFrance to report, “The Amazon Button Is Real (What even?).” But real it is, and according to Daniel Rausch, Director of Amazon Devices, it’s successful too.

What “successful” means at the world’s biggest retailer is complicated. Rausch won’t reveal sales numbers, but Dash Buttons aren’t exactly sold, anyway. For starters, to use them you need to subscribe to Amazon Prime, which costs $99 per year. Then you shell out $4.99 for each Dash button (every brand requires its own button), but the device comes with a $4.99 credit toward your first order. Amazon is no stranger to loss leaders, and it probably makes enough profit from Prime memberships to subsidize the cost of the devices anyway.

But the real aim of Dash Button isn’t to sell Dash Buttons. Like Prime itself, the program’s goal is to encourage more frequent and more regular retail purchases. In that respect, Dash Button joins a family of related programs, including Prime Pantry, a grocery essentials service; Subscribe and Save, a subscription staples service; Amazon Echo, a home-automation appliance that takes orders by voice; and Prime Now, a same-day delivery app available in select markets. On its own, Dash Button seems ridiculous. But as a tentacle swelling from the increasingly pervasive retail monster that is Amazon, it starts to make sense.

“Customers should never run out of stuff,” Rausch implores. If 1-Click and Prime gave Amazon customers the convenience of quick purchase and quick delivery, respectively, then Dash Button hopes to distribute that convenience and speed throughout the home.

Those of us who are shackled to our smartphones might wonder: Why not just use the Amazon app to order a new carton of Huggies or box of Larabars? But from this perspective 1-Click doesn’t make sense either—the shopping-cart checkout process isn’t so laborious, after all; consumers just become less tolerant of any inconvenience as new conveniences arise. And Amazon has built its empire on incremental shifts in retail convenience—starting with e-commerce itself. Dash Button makes certain kinds of purchases ever more convenient by moving the act of buying something slightly closer to the moment when the need for it arises. It’s the famous Amazon 1-Click button made flesh.

But what about my daughter’s sneer at the two-day wait? Rausch tells me that Amazon doesn’t see Dash Button as an impulse purchasing gadget, but a means to keep staples in stock without the nuisance of storing bulk purchases, remembering to re-order, or being tied to a subscription schedule. “Customers know the about-to-run-out-of-it moment,” he explains, citing the bathroom as an example. As you stand before the fogged mirror in your towel and deploy the last razorblade, depressing the nearby Gillette button becomes an act of planning rather than of desperation.

Even so, Dash Button is still fairly limited. Because the devices are branded, Amazon has had to negotiate arrangements with each vendor to participate in the program. And even the few vendors on the roster don’t allow their whole catalog to be purchased via Dash. After shelling out the five bucks for a Gatorade button, I was disappointed to learn that I could only order flavor variety packs—cases of just the non-ghastly red flavor are only available via Amazon Pantry. And so, a phalanx of yellow Gatorades borders my fridge. That’s Rausch’s about-to-run-out-of-it moment, but contorted, as if Amazon knows that everyone hates yellow Gatorade but tricked me into compliance since I could still swill the stuff while waiting for a new shipment of red and orange. Once only yellows remain, it’s time to press the button again.

Dash Button doesn’t just move purchases closer to the moment of need. It also helps produce and cultivate needs where they might never have existed. Amazon could have made a plain button capable of ordering any Prime-eligible product (“It’s a great idea. We’re just not there yet,” they told me). Instead, it negotiated partnerships for branded buttons tied to specific manufacturers. In so doing, Amazon leverages the power of brand identity to “inspire” just-in-time purchases. Oh, Huggies, the Dash Buttoner remembers on the way out of the nursery. Click.

If Amazon.com ushered in e-commerce, Dash Button and its cousins inaugurate a new era of ether-commerce. This is commerce sublimated into the secret recesses of ordinary life rather than commerce conveniently accessible from the computer. And not digital commerce, either, the comparatively easy domain of digital downloads and in-app purchases. Ether-commerce winds its way around you, invisible, like an H.P. Lovecraft monster. And then two days later boxes of durable goods show up at your door, along with the associated Lovecraftian chills.

The ultimate endpoint of ether-commerce is full automation: purchasing things without even really buying them. Amazon’s already in that business, too. In late 2015, its Dash Replenishment Service (DRS) came online. It’s Dash Button but without the button: Manufacturers large and small can bake automated purchasing into their devices, so long as those devices run software and can talk to the Internet. As of January, Brother and Samsung printers, Gmate glucose meters, and GE washers began ordering supplies for themselves without human intervention. And this week, Amazon announced a new partner, the Brita water filtration system. A new wi-fi-enabled Brita “smart pitcher” keeps track of how much water has traversed the filter and automatically orders a replacement when needed.

It sure sounds convenient. But letting your devices decide when they “need” more supplies puts a good deal—perhaps too much—trust in manufacturer honesty and accuracy, as anyone who has had to bypass a printer’s premature “toner low” message can attest. Consumer goods like razors, printers, and water-filtration devices are often sold at steep discounts (or even at a loss) so that companies can reap later profits from expensive supplies. Once the devices automatically place orders for materials deemed low, consumers may lose even more control than they already have over the cost, waste, and freight impacts of their purchases.

Amazon takes an agnostic position on the matter. All the Dash Replenishment Service-enabled device does is order the supply, Rausch insists. Ultimately, the customer still decides when to replace the filter or cartridge, but at least one will be waiting on the shelf when it’s needed. A fair point, but also a convenient way for the retailer to extract itself from debates about the ethics of automated razor-and-razorblade businesses while still making money selling wares on their behalf.

But for now, at least, the pleasure of button-pressing remains a part of the e-commerce shopping experience. Dash Button is designed so that it allows only one order until the last one has shipped, so that overly-eager kids don’t mash the button and flood the porch with Cottonelle. Or adults: I press the button liberally after each order for no reason other than to push it without consequence. And some customers have even trained their dogs to press the button to reorder pet food—a Dachshund Replenishment Service to bridge the gap until Dash Replenishment Service goes mainstream.

The whole thing seems preposterous, but didn’t e-commerce once seem unlikely, too? In time, inevitably, we’ll adjust to robot shoppers, for good or for ill.

It’s now some months after my first Dash button press, and I’m reading on the couch when my phone buzzes. It’s an Amazon notification—an order of Gatorade has been placed. My wife had just gone to get something from the downstairs fridge, on whose door the Gatorade Dash Button is mounted. I tap a few times on the phone, just in time for her to alight from the basement.

“Hey, you accidentally pressed the Dash Button,” I tell her. “I guess it’s easy to do with it there on the fridge door.”

“No, I pressed it on purpose. We were running low on Gatorade,” she responds.

“Oh.” It never occurred to me that we would place an order deliberately, the way the button intends us to. Up until now Dash Button orders have been a spectator event, a curiosity. Eventually, maybe, the chill of Amazon Lovecraft will disperse. But not today.

“I already canceled it,” I admit. We both shrug, and the walls hum with the machinery of houses: furnaces and compact fluorescent light bulbs and refrigerators. Soon enough, they will also purr with the murmur of quiet, alien shoppers, spending my money without me.

Ian Bogost is a contributing writer at The Atlantic.