Amazon Go Get ‘Em? Not So Fast

Amazon Go, the (mostly) people-less, technology-rich, so-called store of the future, opened its doors to the public yesterday.

Advance previews of the space noted both the minimal staff and the heavy presence of futuristic technology: the subway-like gates at the entrance, hundreds of cameras overhead, goods that are automatically added to shoppers’ virtual baskets. But the persistent question everyone’s asking is, Could this model replace traditional retail?

In a word, no.

Here’s what I see stemming from the public launch of Amazon Go.

More technology-rich stores might quickly proliferate. Amazon gets all the credit for developing a store run nearly entirely by technology; it’s an incredible feat. But once someone blows over the physical and technical hurdles of an impossible task, they’ve smashed the mental barriers holding everyone else back too. The impossible is now possible. It emboldens people. Consider this: in 1954, Roger Bannister was the first person to break the four-minute mile but barely a year later, someone else did...and more quickly followed suit. It’s, if Bannister can run it, so can we. It’s, if Amazon can build it, why not us? Now that Amazon’s obliterated the roadblocks, others might be inspired to take their own test drive.

But don’t expect most of retail to follow suit. Let’s be clear: this model is simply not workable for all retailers. Humans are social creatures. We might want every store - from Macy’s to Walgreens - to make it easier and faster to check out (no one, repeat, no one likes waiting in line). But we do generally value human interaction, especially in the Nordstroms, Barnes and Nobles, and Targets of the world. It’s helpful to have people point us in the right direction for products, suggest alternatives, and confirm something looks good (or not). This is why you see Apple, Best Buy, and even Dollar General investing in people, not technology for technology’s sake. Retailers who remember the human side of the equation will ultimately be successful even in a post-mobile, post-Amazon world.

That said, there might be one exception where similar stores could work. Think highly commoditized goods - the stores with very predictable inventory with regular turnover. Convenience stores are great examples. Human interaction isn’t as needed there (no one needs someone to ask if something is flattering). The Amazon Go model is fairly perfect because - like convenience stores (and Amazon.com itself) - it’s super transactional. If you’re running in to a 7-11 to grab Advil or diapers, you’re not looking to linger and the retailer’s not overly focused on getting you to stick around. The value prop here is get in/get out as efficiently as possible. The premium is on product availability and ease of checkout, not the availability of highly knowledgeable, helpful sales associates.

And don’t get excited about Amazon selling this technology to others. The Times had an interesting aside in its launch coverage: "There’s even speculation that Amazon could sell the system to other retailers, much as it sells its cloud computing services to other companies."

I doubt it. It would have to work for both Amazon and other retailers - and that would be very tricky to successfully pull off. Retailers know that they’d probably have to give Amazon their data in order to get access to the technology and setup - and Amazon’s in the business of eating their data for lunch. Safeway, for example, should never consent to using Amazon’s app as part of a similar setup. It’s like visiting a loan shark - solves your short-term problems but in the long run, you are incredibly hosed. There must be a decoupling of the hardware and the software for this to work.

Plus, let’s not forget - this is first generation technology, and calling it that is generous. After paying for maintenance, upgrades, tech support, plus the minimal staffing the store still requires, it may be just as cheap for retailers to stick with the tried-and-true.

Maybe there’s a scenario in which Amazon agrees to allow retailers to use their own apps that get people in the door and handle the transactions. For a fee, Amazon would perhaps outfit the stores with the technology. But what’s compelling about that scenario for them? Unless they offer a reduced fee in exchange for - you guessed it, data - there’s little incentive for them to sell the model.

Ultimately, time will tell how this shakes out. For now, I’m waiting to see what happens.