Hacker News Reader

Comment by CydeWeys | Hacker News Reader

Isn't the whole point of Amazon generally, and these stores specifically in not having staff, that it's supposed to be cheaper? I think you've pinpointed the reason it failed, but I can't imagine why they thought it was going to succeed at those prices.

View on HN

How did it fail? They just announced that eight stores were permanently closed after having been closed indefinitely, one of which was inside of an Amazon office that shut down last year.[1] It looks like they still have 29 Go stores open nationwide.[2]

[1]: https://www.geekwire.com/2023/amazon-closing-eight-amazon-go...

[2]: https://www.amazon.com/b?node=16008589011

If you have a US-wide convenience store chain with a multi-billion-dollar company behind it, and you have 29 stores after six years, well... I mean, it's not exactly a stand-out success, is it?

I assumed that every one of those stores were experiments or at least test jigs, and closing or opening them means nothing to anyone outside Amazon.

NYC is a tough market because there is so much competition. But if you're not trying to make money yet you can learn a lot. Perhaps they learnt that New Yorkers don't have the patience to install an app. You generally don't wait that long in a bodega -- I expect to wait longer inside a gas station on the Interstate!

well, Amazon Go isn't the only cashierless brand in Amazon's portfolio. It was the experimental small store. They're moving towards deploying the technology in Whole Foods and Amazon Fresh stores.

This reminds me that the San Francisco ones, if I remember correctly, are in a part of town that's been famously devoid of foot traffic since the pandemic.

San Francisco might be receptive to the concept, but it would have to be close to where people live, rather than where office workers used to go.

The only one I know in SF was down in the FiDi, and I can’t imagine that area had much traffic during the pandemic, and I question how much it has now.

That's precisely what i was saying.

I'm also seeing this tech in some airports in basically a linear contained store. Walk in, walk down the tunnel grab what you want, walk out. Seemed very efficient and airport appropriate

Sports venues, too. They have the Amazon version installed at Climate Pledge Arena and a few baseball stadiums, but I've seen something similar at a walk-up beer fridge during a Nets game at the Barclays Center.

Just because they're blowing money on it doesn't mean it's a successful concept

My hunch is that the store wasn't meant to be particularly popular with lots of customers, it was meant to demonstrate technologies they wanted to sell to businesses.

https://aws.amazon.com/just-walk-out/

https://aws.amazon.com/one/

They've employed a "Amazon is the first and best customer" for most of the services they've launched. Amazon.com / AWS is an example.

Was Amazon actually the first customer for AWS (being a very early customer counts)? I've heard that they were, but I've also heard that the AWS guys designed for a different customer base first.

Like the eBay Pez story it makes sense and adds credibility for the narrative to be that Amazon was the first customer for AWS. But it feels like a very unamazon thing to do, given that so much of its customer base isn't running the kind of enterprise jobs that an Amazon scale company needs.

(Either way I imagine almost all of Amazon does run on AWS these days and I'm sure the transition took a long time)

iirc aws exists mainly because amazon kept needing capacity, and after solving that problem it then had a bunch of excess capacity

Excess capacity story is a myth. Frequently corrected/debunked by Werner Vogels.

I'm not sure it actually failed. These stores work by having AI track customer actions via camera, and over the last couple of years they've gotten a ton of data and domain knowledge from them. It isn't hard to imagine that's been leveraged in whatever's watching the cameras at Whole Foods.

The point of spending this massive amount of money on an automated store is to recoup that investment by opening lots of automated stores and make money in volume, taking advantage of your lower labor costs.

If they don't actually open at least hundreds of these then they've thrown away massive amounts of money on R&D they'll never get back, and the data they did accumulate isn't even particularly useful as it's in the context of a type of store they're not running at scale.

Although they've added self-checkout lanes to Whole Foods, human interaction is in its DNA and you pay for that premium.

Amazon is only able to capture your order details if you use the app at checkout.

What kind of human interaction do you get out of some bored 15 year old girl at the supermarket? For me it doesn't go beyond "hello" and "bye". Maybe it's something to do with culture.

Anyway I looked up some data and over half the customers in my country use the self checkout now.

It could have been intentionally set to a more expensive price level, to manage the number of people using it. (To manage the "load" on the product development team, because if 2% of all users encounter some problems - or otherwise need support - then it's not scalable.)

I wouldn’t be surprised if that were the long-term goal but in reality it was not yet cheaper. Automation often proves more difficult than one would expect.

https://xkcd.com/1319/

This should be branded on the forehead of every CEO of a company that promises full self driving capabilities for cars before 2030 and cryptocurrency mass adoption, also before 2030.

The image caption text seems very fitting as well. Not quite the dictionary definition of auto-mation but maybe it should be.

Cheaper for Amazon, yes. Why would they externalize the profit? Just maximizing extraction by eliminating the few remaining local jobs.

They also recently opened cashierless Whole Food locations that use the same technology. I would not be surprised if some of these Gp stores were never intended to stay open but rather existed as a concept and to test ideas.

Maybe they were thinking about the immediate gratification cognitive bias. You might see the price on the shelf as you take each item from the shelf, but that doesn't translate to a total expense until you actually pay the Amazon bill at the end of the month.

> Isn't the whole point of Amazon generally, and these stores specifically in not having staff, that it's supposed to be cheaper?

I'm pretty sure the goal is to have a higher profit margin than competitors by leveraging technology, not lower prices.

Reply on news.ycombinator.com