Amazon Go and why 'Frictionless' shouldn’t mean Humanless

Perry Hewitt, Senior Advisor at Lincoln Center for the Performing Arts, reviews Amazon Go’s ‘just walk out’ shopping experience first-hand, while Lindsay Herbert, Inventor of IBM’s Instant Checkout, argues that giving jobs to technology in stores shouldn’t mean taking the human jobs out.

 Frictionless in Seattle: Embracing the Panopticon of Amazon Go

By Perry Hewitt

On the final evening of its first week open to the public, the Amazon Go store still drew lines of eager customers. The lines were staffed by employees in orange parkas, who cheerfully engaged with shoppers and handed out high-quality, reusable bags. And the wait was short: no more than five minutes in the Seattle drizzle.

As a frequent Amazon shopper, I experienced a kind of brand disconnect when physically confronted with Amazon mark on something the size of a convenience store. Living in New York City without a car, I use the Amazon apps on my phone far more frequently than I ever drove to the grocery or hardware store back in Massachusetts. While Amazon apps give me a wide selection from workout gear to floor lamps, the physical Amazon Go store was aimed at grab-and-go behaviors: quick snacks, prepared foods, and even a mini liquor store. It also took a moment to re-frame the experience in my head: In this physical Amazon store, the stock is limited, and can actually run out.

The main event, then, is not the stock selection but the technology. Entry is easy: Anyone who has scanned a mobile boarding pass in an airport would find the turnstiles familiar. In the store itself, there is nothing to scan or interact with to check price or, of course, check out. This is a sharp contrast to the typical, painful automated checkout experience, where there is usually an exhausted staff member assisting the general public in completing each “self checkout” transaction. At Amazon Go the cameras above, and there look to be thousands, are doing the counting for you. And when you leave, you walk out -- a behavior some find unsettling, but that I believe will be as easy to adopt as a one-click purchase on an app. 

And the technology is breathtaking: I purposefully picked up and replaced more than 10 other items, to see if one would be mistakenly added to my cart, but it was impossible to fool the system. I overheard some shoppers saying that the checkout had experienced issues earlier with the “cameras having trouble reading body type” and charging to the wrong person, but saw no evidence of this. Looking up was instructive: seeing those cameras was a reminder that some of us have walked willingly into the panopticon, trading privacy for seamless commerce. This feels also like a first step: the technology has removed the weight of the interaction and will soon deliver more intelligence about our products and their provenance.

And what of the people? There were several staff at the front, assisting the very occasional shopper who still needed to download the Amazon Go app or had trouble scanning the barcode to get in. There were two people intermittently re-stocking or adjusting product, another back in an employees-only area, and a rather gruff man checking ID in the mini-liquor aisle. The employees were largely friendly, and checked in frequently with shoppers to see if assistance was needed. Oddly, when I asked one what he did he replied, rather ominously, that he could not disclose what his role was. Note to Amazon comms team: “customer support” is a broad and useful catch-all phrase.

How will all this work out for Amazon? Ben Thompson makes a compelling case for Amazon’s thus far successful strategy of pursuing both a vertical and horizontal model. What it means for other kinds of transactions beyond in-store purchase is perhaps more exciting. If a technology can allow you to know the people in your physical space, be it a sporting event or a concert venue, and eliminate bottlenecks, the potential for improving audience experience is enormous.

Rise of the Robots vs. Resurrection of the Shopkeepers

By Lindsay Herbert

As the inventor of IBM’s new instant checkout technology, I’m often drawn into conversations by people who think that any new technology is designed to eliminate the need for human workers.

It usually starts with “but I like chatting to people in stores,” followed by “I suppose it’s the future though – pretty soon everything will be run by robots.”

The reality is that technology exists to serve humans – freeing them from tedious tasks so they can apply their creative and adaptive minds to bigger pursuits. That means the job of retail technology should be to elevate the role of the shop assistants – not replace them.

I happen to think that a store without human staff would be a poor excuse for a shopping experience – like browsing inside a vending machine. I live in central London, where online shopping can be delivered fast and for free, so when I go to a store, it’s because I don’t know what I want yet and I’d like some inspiration or guidance. 

Artificial Intelligence, sensors and cameras technology should be used to give insights to retail staff, so that shopping in any store could feel like the old days when shopkeepers knew their customers, had deep knowledge of their stock, and could foresee trends or scarcities poised to impact supply and demand.

However, if workers of tomorrow are going to have to interact and extract value from an Artificial Intelligence, it means a dramatic shift from the workforce expectations of today that for some, extend only as far as pressing buttons on a till, or balancing numbers on a spreadsheet.

Technology companies are racing to perfect intelligent machines, but what are employers, educators, and governments doing to prepare people for roles that involve working with it?

In my book Digital Transformation, I share insights from experts worldwide and evidence that the pace of change today necessitates new ways of working brought about by the innovation age.

Instead of waiting for mass unemployment to force a change, companies should be trialling these new ways of working now. This includes a move away from traditional company structures with siloed departments and top-down hierarchies that restrict the flow of information, and reduce opportunities to trial new solutions.

Retailers, give your store staff problems to tackle collaboratively, and give them the freedom and means to try and test new solutions and technology iteratively. When it comes to daily operations, help your staff establish feedback loops with customers and other sources of data that they feel empowered to act on – this will help them, and your brand, to anticipate and react to changes and opportunities. This style of management can help workers of all kinds begin to adapt to new technologies incrementally, while the companies that employ it get to leverage the best of what’s available now.

Technologies that mean I can ‘just walk out’ with my shopping shouldn’t mean I don’t also get to have a great chat with a well informed local shopkeeper. Instead, use the technology to take me back to that golden era of retail when the people behind the counters were the masters of their domain – and don’t tell me there’s no place for human jobs in the future of retail when it’s humans that the technology was invented to serve in the first place.