Just curious if we have any statistics about how much time is saved per consumer?
TBH I don't feel particularly happy about programmers removing retail jobs. Yeah for sure technology advances always remove some people from their jobs, but there are consequences.
Not sure about data on time saved, but I average 2-5 minutes on each trip to the Go store in Seattle.
As for removing retail jobs, Amazon does hire a ton of people for every store to restock, answer questions, prepare deli foods, etc. So the jobs are likely moving to different parts of the store.
Also, I’m not sure what the official goals of Go are, but I always thought it was to remove friction during checkout. Not to save costs on cashier jobs.
Disclaimer: I don't work for Amazon. They didn't want to hire me.
> I’m not sure what the official goals of Go are but I always thought it was to remove friction during checkout.
I visited the Seattle store one evening in April while in town for an Amazon onsite. It was completely empty from customers but there was a dude sitting at the entrance to the alcohol isle asking for ids and a girl trying to fit some products into some curly thing.
> As for removing retail jobs, Amazon does hire a ton of people for every store to restock, answer questions, prepare deli foods, etc. So the jobs are likely moving to different parts of the store.
Is it fair to say that you are making the claim that there is no net decrease in labor hours when compared to a similar store that does staff cashiers?
Why would that matter to how happy someone feels about it? Just because nobody throws clogs in looms anymore doesn't mean the tech transition was handled well at the time.
I was a pretty regular user of the Go store near my commuter train home.
Lately there are numerous items at my store that have gone out of stock, some of them for a few weeks now, and telling them about it gets a shrug and a "sorry, 'corporate' controls inventory" response. Not sure what 'corporate' means here but okay.
Long story short - Go is a fun thingy but I don't think they're ready for retail. I deleted my app.
Interesting, at my Trader Joe’s they always ask, and when I do say I couldn’t find something they just shrug and apologize. Ironically this is the TJ’s up the hill from the new Go store mentioned.
Trader Joe's is a bad example, they are notorious for discontinuing items that never return to shelves. This behavior is referenced in this old viral video https://www.youtube.com/watch?v=OdB7GDZY3Pk
That was meant as an eye roll, as in “if Amazon Seattle already considers this high-visibility project as being an arms-length away, then perhaps it’s not as important as they make it”
Ugh this has happened to Whole Foods too. Evidently they started making the change before Amazon but it's getting worse. I'm still bitter they closed my neighborhood store, but when I make the trek to the next closest one only to find a handful of the items I regularly get are out of stock is frustrating. And it's usually not perishables which I can understand.
This early, they’re probably running costumer/data experiments on out of stock scenarios.
How long can a thing be out of stock before its sales dip upon return? Do users who regularly buy those things stop visiting? Buy other products in the same category? (Do they switch back when stock resumes?). Etc...
Airports make a lot of sense to me. Movie theatres kind of do. Supermarkets don’t. If I squint hard enough I can see it working (I’ve been to the Go store a couple times and it is indeed a great if weird experience), but food like vegetables seems harder to do this with than prepackaged sandwiches and sodas. How do you handle things that are priced by weight? For example potatoes, or steak. Make the customer print a label with a QR code the cameras can read? That sounds like a terrible experience, and the “solve” is more packaging and trash - something Amazon already causes a lot of.
Here in Germany, it's common to weigh goods that are priced by weight and then have a sticker printed out to stick on them for the cashier. I don't see printing a QR code or whatever weird mechanism Amazon uses to price items being much more complex
Interesting. I am skeptical that US consumers would have the discipline to do this. Maybe another way to solve this problem would be adding a weigh station next to each SKU (or group of SKUs), and then the station logs the weight in the app as the customer puts the item in their bag?
Depends on the state in the US. Almost every state has relatively strict requirements regarding which models of scales may be used for trade (pricing), and some impose taxes and other fees to keep them in operation. Most scales on grocer floors that are accessible to customers (if you can even find them these days) are clearly marked as "not legal for trade" or have a similar disclaimer -- precisely because it would be too expensive to keep carefully calibrated trade-legal scales exposed to the general public.
Meijer (US megastore retailer that competes against Walmart in the American midwest, though generally has slightly higher prices and higher quality items) has introduced this recently with app-based self-checkout, and it's picking up adoption at a fairly quick pace.
Another huge regional grocer, HEB, does this in Texas with their largest and most upscale stores, and it's been a norm for nearly a decade.
Forcing the customer to drop the item(s) on a scale seems most reasonable, but requires relatively strong controls to ensure the customer isn't pulling shenanigans like placing item A and item B into a bag, then co-mingling them when weighing them later.
As long as you don’t have more than one person picking up produce from a bin at the same time you can just measure the change in weight from the bin it’s taken out of. That has to be super cheap technology at this point.
Seems complicated. Even if you're resetting tare on the bin regularly, fruits and vegetables can have significant changes in weight due to dehydration, and most markets are doing things like misting chilled produce with water, which will introduce constant and somewhat unpredictable changes in bin weight.
The tare would be reset each time a customer picks something up. The total bin weight doesn’t matter, just the delta between before and after the computer vision system identifies that a customer removed some of the bin’s contents.
>The tare would be reset each time a customer picks something up.
This would run afoul of commerce laws in pretty much every US state. Tare must be set from zero, not relative to whatever happens to be on the scale, and certainly not on the fly.
> How do you handle things that are priced by weight?
I assume they'll handle it the same way Amazon Prime Now seems to handle delivery from Whole Foods: it's a platform where weight doesn't actually matter.
I get insanely variable results when I have Whole Foods items delivered via Amazon Prime. 2 lbs of chicken breast might be as little as 1.2 lbs, or as much as 2.5 lbs. In either case, it's priced as 2.0 lbs. If you're savvy enough to weigh it and realize you've been stiffed, Amazon refunds the entire item, so it's a win for the consumer...I guess?
In general I find that my weight-based fruit/vegetable/meat/bulk bin item purchases via Prime Now slightly exceed whatever I actually requested and paid for, and Amazon never charges for delivering more than you initially requested. But maybe 1/8th of weight-based items will be drastically under-weight. I have a kitchen scale handy, so I always catch this and report it. Amazon has never failed to refund the full cost of the item, but I do wonder how many other consumers are never noticing that they've been scammed when Prime Now fails to deliver the weight that was paid for.
For produce, you could turn the shelf into a scale. It reads X before the camera detects the customer picking up the item to put in their cart, and it reads Y after. Therefore the potatoes they picked up weigh X-Y.
Another approach would be to have a scale and require them to stick their potatoes on it, but not require the other, tedious steps like entering a PLU code or printing a sticker. If you can track them picking up items off a shelf and putting them in their cart, presumably you can also track an additional intermediate step where they put it on a scale. (And connect that scale to the network, take a reading when you see them stick potatoes on it, and correlate that.)
Of course you can always just sell by unit and not in bulk. Sell 1 pound, 2 pound, and 5 pound, and 10 pound bags of potatoes instead. Maybe even 0.5 pound if someone wants a small quantity. Or for things whose size doesn't vary much (like apples), just sell by number of apples.
For steak, the packages are normally already weighed and labeled, so the problem might be easier if you can just scan a label that's already printed.
As an European, Amazon Go seems like extreme over-engineering, with its vision-based system and all. I can already quickly scan items as I put them in my basket, either using a handheld scanner you pick up at the entrance, or my own mobile phone. You can also do it all at once in the self-checkout station. It already takes little to no time, and simply adding a very fast barcode reader to the basket itself or RFID would completely eliminate any time spent there. Payment takes five seconds with contactless / Apple Pay. Most small supermarkets already have only one or two cashiers working at any time.
Vision system replaces cashier and loss prevention. Plus I imagine just grabbing stuff lowers the barrier to sale, versus having to scan each item and acknowledge the cost.
This was how I felt about Amazon Go...until I tried Amazon Go.
I had previously tried other markets with handheld UPC scanners or mobile device scanning, but still ended up with delays I'd rather avoid (passing through the queue at self-checkout). Amazon Go really is completely effortless once you have checked in to the store, although doing things like putting items into your own backpack while shopping feels...very weird. But being able to waltz out the door with zero delays is a true luxury in urban areas where markets almost always have long queues for checkout (regardless of if they have a human attendant or not).
I’ve been only a dozen times, but the experience gets fermented in your mind such that you will accidentally try to walk out of a 7-11 without paying. It really is effortless and a totally new customer experience. I’m still not sure if it’s a net positive, but just scanning in and walking out is great.
My local (US) grocery story has a similar scan-as-you-go feature available, and I haven't seen a single person using it in the year or so that it has been available. For older folks and the tech-illiterate, it is likely somewhat intimidating to have to teach yourself a new 'techie' feature on the fly.
Amazon's method may require you to log in or scan in with a barcode from the Amazon app, but just about everyone already has some basic experience with the Amazon service. And once you do a single scan, it just works.
There is an immense difference between what you describe and the user experience of Amazon Go. It was one of the most revolutionary experiences once I got accustomed to it.
I used to regularly get lunch in < 30 seconds at Amazon Go. My record was 12 seconds for a soda and sandwhich. They added more processing time for that calculation at some point. But I basically could get things as quick as my feet could take me in and out.
It's really not the same experience. You should try it when you get the chance.
Here are my most recent trip times according to the app (assume a 15s overestimate to my real experience):
37s - Dr Pepper
32s - Dr Pepper
1m12s - Dr Pepper, sandwhich, cookies
33s - Dr Pepper
1m33s - Dr Pepper
36s - Dr Pepper
44s - Vitamin Water
46s - Sandwhich, cookies
30s - Vitamin Water
32s - Dr Pepper
31s - Dr Pepper
35s - Dr Pepper
I would estimate that most of these have a 15s overestimate.
I can and have bought a Coke Zero in that amount of time by using self-checkouts with Apple Pay in Sweden (specifically the ones at Pressbyrån and 7-Eleven). Walk in, grab bottle, scan and press pay, Apple Pay Touch ID, a few seconds for the transaction to approve, and I'm out of the store.
Are there never lines at the checkout in Sweden? Here in the US, we have self-checkout, and we have contactless payment too (though not my stubborn grocery store), but you still often have to wait because the process is slow and the store gets busy.
Though I've never tried a Swedish self-checkout system I'm fairly convinced that it can't possibly be as fast as the Amazon Go stores. However adept you are at manipulating the self checkout machine and paying, that can't compare to the greater speed of not doing that at all and just walking out the door.
Shopping at the Go Store is much faster. When you're grabbing a single item you can have times of ~10 seconds. Grabbing more items and you save more time by not having to check out. No checkout, no lines, no figuring out why your phone or card isn't accepted, or what error message is on screen etc. In and out as simply and efficiently as possible.
I wouldn't say it is revolutionary, but it does cut down the time it takes to go shopping. It's like a perfect version of self-checkout.
I have bought single things using self-checkout that quickly, but I think the difference would be when you buy 5 or more things. Then the time to scan becomes noticeable.
Another, more subtle feature is that I never wait in line because I have trained myself to never go at peak times. I waste a certain amount of time thinking about it.
That said, I find the idea of a cashierless store super-creepy and will be among the last adopters.
Yeah these times seem really high. The app reported that I averaged about 12s for grabbing the same item every morning (boiled eggs) for about 6 months.
And I think you are overestimating the amount of work in the process I described (hence, overengineering). You can totally grab a soda and walk out in <30s here. Supermarkets are local and small, you’ll rarely find a line at self-checkout.
These methods work by turning the shopper into a cashier. Amazon go completely removes the cashier role from the experience, the shopper just picks up the item as they did before and walks out.
Using vision to solve a problem always seems inefficient but it scales out way better than anything else (ie the various ways google maps uses vision when in a world with perfect map data it wouldn’t be needed).
I had those problems in the past but I don’t think I’ve had a problem with self checkout in the last couple years. Granted, all my experiences are at Walmart so maybe they have figured it out. In addition, the whole process is quick and painless. I scan a QR code with the Walmart app when I get to the self checkout, scan my items, press done, and leave.
Also, the Scan and Go system they have at Sam’s Club is awesome. I can scan items on my phone as I’m shopping. When I’m done, I tap purchase, and then walk out with my stuff.
Wow if they can convince their competition to use the system, and they can use that data, that would be hugely valuable. Not just for grocery/online sales but I'm thinking of data use for their fast growing ad network.
I've enjoyed the convenience of the Go stores since they opened, but one thing that's popped into my head is how 'classist' this type of store is, especially in a city like Seattle where they started.
To be allowed into the store requires a few prerequisites which unfortunately not everyone has the benefit of possessing; smart phone with data service, credit card, amazon account.
For most people this is no problem. But what about all the people who rely on cash based systems to exist? What about people who don't have smart phones with internet plans (though this is becoming an increasingly small population, it's still very prevalent)?
Now what happens to those people when this becomes a standard and the alternative options diminish? Food deserts in underserved (poor) neighborhoods are already wide spread in the US, this credit based system which relies on an app will only make it harder for people who already suffer.
There's an easy answer that's already prevalent in markets like Japan - just accept the local transit card as a payment option. It's something you can obtain with almost no regulatory friction whatsoever, and top off with cash easily. Likewise, it's treated like a cash payment at quick service restaurants, grocers, convenience stores, and many other types of retailer.
The one problem Amazon would have here is preventing customers from leaving if their purchase exceeds their cash balance on the card. Seems relatively easy to stop with a 2-step system for exiting (cross a line/barrier to signal that you're leaving, wait for a physical barrier to open after the payment method is debited).
It doesn't strictly have to be that. The prepaid/debit/credit/charge card w/ NFC one uses to enter the premise can be the one that's charged against when one leaves. The card issuer can notify you of the charges once Amazon posts them. Amazon can verify for card-fraud as they already have tracking hardware installed.
You still have to have a membership unless you have a Costco cash card or go with a friend, and $60 or whatever upfront can be too much for many people. Also, it's a bit more expensive per visit since you need to buy in bulk.
That being said, we have plenty of grocery stores in my area and they seem to compete quite well with Costco. I am not worried about these stores, and I don't plan on using them because I find them to be a privacy issue. I think there are plenty of people who either prefer cash or don't want to tie their food purchases to their Amazon account to keep traditional grocery stores in business.
I think you are being too optimistic with the last sentence. Whenever there is a tangible benefit that only has "mostly unseen" downsites, such as food prices being low since you're subsidizing Amazon training their rekognition software, people will take that benefit since the downside doesn't mean much to them (at least in the moment). Most people see their grocery bill going down non-unsubstantial amounts and take that as a good tradeoff for having cameras and sensors figure out what you're putting in your bag, and then later seeing that grocery visit in your Amazon order history.
On HN we often hear there are plenty of folks who like paying cash and don't want to tie purchases to a corp system.
Then I go to a place like safeway and literally everyone pays with a credit card and gives up their phone number so they can get MAJOR discounts on purchases (ie, $1/item vs $3/item) and points towards things. Even the folks on EBT are using some kind of credit cards instead of the old vouchers.
Can Amazon Go stores take EBT cards (aka food stamps)? I imagine not… even if they could process the payments, how would they enforce the restrictions on what food stamp recipients are allowed to buy?
The same thing happened when commercial airplanes started, only the rich could afford tickets at the beginning, later they became affordable, so it's possible something similar well happen with automated stores.
Like with most things, poor people make up a bit enough portion of the population that stores will always cater to them. Also, plenty of people value their privacy more than convenience, and those people won't shop at these Amazon stores.
Just look at Amazon.com vs Walmart, Walmart is still in business, even though running a brick and mortar store is more expensive than an online store.
Online is not always a net gain for the retailer. For something like 24 packs of water at Walmart it’s simpler to ship several pallets to one location and have people so last mile delivery to their house than to ship it directly to them. It’s largely shoplifting and floorspace that’s in online stores favor, but that’s only part of the equation.
You can enter without a phone and pay for items with cash. I found this information on the internet. But feel free to have a 20 deep discussion about it.
I think the answer is not to hold back progress, but to address the pain points as they arise. As for the specific concerns of going "cashless", there are many countries with lower standard of living (Venezuela, China) than the US but near-ubiquitous cellphone payments, so I don't see this as a difficult problem to solve.
Amazon accounts are free. Surely you must complain much more vocally about Costco, Sam's Club, and other retailers that literally charge you money for the rights to shop there, right?
Amazon Go stores are located in very dense urban areas and—if successful—will drive out the 7-11s and other convenience stores in the area that accept cash.
Costco, Sam's Club, and other similar 'big box' stores are usually located in less-dense areas and are not competing directly with 7-11s or other convenience stores.
It’s not hard to get a cheap smartphone (something like 60-80% of homeless people own a smartphone based on a quick google) and if you qualify for SNAP, you can almost certainly get a free smartphone through the lifeline program.
> data service
Amazon Go stores have free WiFi.
> credit card
Edit: Originally stated you can use gift card credit, but that’s incorrect. You can get a secured card with no credit check though.
> amazon account
Free and obviously easy to get if you have access to the internet, which virtually everyone does through a public library.
It's usually only obligatory to accept cash for settling a debt, e.g. if you have eaten in a restaurant, the restaurant is obliged to accept cash, because you can hardly give back the food you already ate (at least in the same form). But if you are at McDonalds, there's no obligation for them to accept cash because you pay upfront.
I haven't stepped into a Amazon Go store yet, but the tradeoffs in requiring an account/app have been on my mind after experiencing the new Seattle Target in Ballard. It's entirely self-checkout (with an assistant or two there), but the dystopian part during the self-checkout process is an LCD monitor and camera right above you showing your face with "RECORDING IN PROGRESS" above.
The Amazon Go convenience store at 920 5th Avenue (looking from the outside) felt a bit sad in their dining setup - it's a tiny bar that seemed designed for quick, alone consumption. I'm guessing most folks going there grab a lunch then go back to their office to eat.
(image: https://www.google.com/maps/place/Amazon+Go/@47.6063189,-122... )
Given the makeup and general behavior of the Seattle population you describe, I’m quite confident deterring them from visiting was absolutely intentional by Amazon
Raising the minimum wage to $15 and higher will hasten cashiers being replaced by machines. In the unionized grocery store I patronize, where there was recently a strike, they have reduced the numbers of cashiers and increase the number of self-checkout stations.
In sympathy with labor, my wife refuses to use self checkout systems. She says they're not paying her to do that job. When store managers or employees attempt to direct her to the automat, she refuses and lets them know why. They've told her she isn't the only one who prefers humans at the check out.
I think it's fair to say they are paying her to do the bagging in the form of "lower prices" as a result of paying for less employees, but I get the concern.
Walmart has done this near where I live. Previously where there was maybe 10-12 manned checkout stations, there are now about 12 or so small self checkout stations and 12 large checkout stations with a lot of room for bags. There are still maybe 10 manned checkouts, but only a few are open at any time so the quickest option is usually the self checkout.
I went to an Amazon Go store for the first time recently. I grabbed some sushi and a wrap, and it was surprisingly tasty and inexpensive. For the location and quality, it made me think they have to be taking a loss (or just breaking even) on the products.
The store itself was nice enough. But it was small, and felt like a novelty, like a lab experiment and not an actual convenience store. As a software engineer, I was also keenly aware of the massive array of tracking technology all around me, and that didn't feel great.
I don't really have any complaints. It was convenient, but not "oh wow this changes my life" convenient. If this kind of technology pushes down the prices (though that's honestly hard for me to see how), that will be a benefit.
I wonder how well these will do with the grab-and-go gangs that are currently working higher end retail in Chicago. Do they have reasonable security or simply depend on cameras to identify perps?
I think I remember seeing (from the initial wave of news) that they tell you not to intentionally try to obfuscate or trick the system, but if you pick up a product and it doesn't correctly track it, it's yours (but you should report it so they can improve tracking).
The busier the Amazon Go store the more employees are on-site.
The one in the West Loop is right next to the trains and sees a lot of traffic during rush hour. So they have 4+ employees just standing around watching people. The ones at random locations in the loop aren't as busy - so only have a couple employees in those stores.
Actually I think it's weird that Amazon has so many employees in Go stores. You'd think they'd just need 1 person to stock things and call the police if someone jumps over the gate without scanning their phone. Every time I see a group of their employees standing around I want to ask "why haven't any of you been replaced by robots yet?"
At my office there’s a fridge with a card reader. You can take whatever you want out of the fridge after swiping your card. There’s breakfast and lunch and snack items in the fridge. Some company has already replicated the Amazon Go model for a use case that ... I would argue ... makes more sense. (Unless they were first)
132 comments
[ 3.6 ms ] story [ 82.5 ms ] threadTBH I don't feel particularly happy about programmers removing retail jobs. Yeah for sure technology advances always remove some people from their jobs, but there are consequences.
Not sure about data on time saved, but I average 2-5 minutes on each trip to the Go store in Seattle.
As for removing retail jobs, Amazon does hire a ton of people for every store to restock, answer questions, prepare deli foods, etc. So the jobs are likely moving to different parts of the store.
Also, I’m not sure what the official goals of Go are, but I always thought it was to remove friction during checkout. Not to save costs on cashier jobs.
> I’m not sure what the official goals of Go are but I always thought it was to remove friction during checkout.
I visited the Seattle store one evening in April while in town for an Amazon onsite. It was completely empty from customers but there was a dude sitting at the entrance to the alcohol isle asking for ids and a girl trying to fit some products into some curly thing.
Friction is (re)moved from point a to point b.
Classic 2010s tech play.
Is it fair to say that you are making the claim that there is no net decrease in labor hours when compared to a similar store that does staff cashiers?
Is this any different than Caterpillar removing ditch diggers’ jobs, or internet remove postal carrier and television channels’ jobs?
If you sit around expecting the rest of the world to take care of your best interests, you’re going to have a bad time.
Lately there are numerous items at my store that have gone out of stock, some of them for a few weeks now, and telling them about it gets a shrug and a "sorry, 'corporate' controls inventory" response. Not sure what 'corporate' means here but okay.
Long story short - Go is a fun thingy but I don't think they're ready for retail. I deleted my app.
They mean someone outside the store manages it. Someone who manages across the corporation rather than just at the retail location.
How long can a thing be out of stock before its sales dip upon return? Do users who regularly buy those things stop visiting? Buy other products in the same category? (Do they switch back when stock resumes?). Etc...
Another huge regional grocer, HEB, does this in Texas with their largest and most upscale stores, and it's been a norm for nearly a decade.
Alternatively, the entire shelf is a scale, but the shelves are smaller so that only one person can pick vegetables at a time.
This would run afoul of commerce laws in pretty much every US state. Tare must be set from zero, not relative to whatever happens to be on the scale, and certainly not on the fly.
I assume they'll handle it the same way Amazon Prime Now seems to handle delivery from Whole Foods: it's a platform where weight doesn't actually matter.
I get insanely variable results when I have Whole Foods items delivered via Amazon Prime. 2 lbs of chicken breast might be as little as 1.2 lbs, or as much as 2.5 lbs. In either case, it's priced as 2.0 lbs. If you're savvy enough to weigh it and realize you've been stiffed, Amazon refunds the entire item, so it's a win for the consumer...I guess?
In general I find that my weight-based fruit/vegetable/meat/bulk bin item purchases via Prime Now slightly exceed whatever I actually requested and paid for, and Amazon never charges for delivering more than you initially requested. But maybe 1/8th of weight-based items will be drastically under-weight. I have a kitchen scale handy, so I always catch this and report it. Amazon has never failed to refund the full cost of the item, but I do wonder how many other consumers are never noticing that they've been scammed when Prime Now fails to deliver the weight that was paid for.
Another approach would be to have a scale and require them to stick their potatoes on it, but not require the other, tedious steps like entering a PLU code or printing a sticker. If you can track them picking up items off a shelf and putting them in their cart, presumably you can also track an additional intermediate step where they put it on a scale. (And connect that scale to the network, take a reading when you see them stick potatoes on it, and correlate that.)
Of course you can always just sell by unit and not in bulk. Sell 1 pound, 2 pound, and 5 pound, and 10 pound bags of potatoes instead. Maybe even 0.5 pound if someone wants a small quantity. Or for things whose size doesn't vary much (like apples), just sell by number of apples.
For steak, the packages are normally already weighed and labeled, so the problem might be easier if you can just scan a label that's already printed.
I had previously tried other markets with handheld UPC scanners or mobile device scanning, but still ended up with delays I'd rather avoid (passing through the queue at self-checkout). Amazon Go really is completely effortless once you have checked in to the store, although doing things like putting items into your own backpack while shopping feels...very weird. But being able to waltz out the door with zero delays is a true luxury in urban areas where markets almost always have long queues for checkout (regardless of if they have a human attendant or not).
Amazon's method may require you to log in or scan in with a barcode from the Amazon app, but just about everyone already has some basic experience with the Amazon service. And once you do a single scan, it just works.
I used to regularly get lunch in < 30 seconds at Amazon Go. My record was 12 seconds for a soda and sandwhich. They added more processing time for that calculation at some point. But I basically could get things as quick as my feet could take me in and out.
It's really not the same experience. You should try it when you get the chance.
Here are my most recent trip times according to the app (assume a 15s overestimate to my real experience):
37s - Dr Pepper
32s - Dr Pepper
1m12s - Dr Pepper, sandwhich, cookies
33s - Dr Pepper
1m33s - Dr Pepper
36s - Dr Pepper
44s - Vitamin Water
46s - Sandwhich, cookies
30s - Vitamin Water
32s - Dr Pepper
31s - Dr Pepper
35s - Dr Pepper
I would estimate that most of these have a 15s overestimate.
This has not been a Dr Pepper advertisement.
It's nice that you save a couple minutes, but I doubt it really has much impact on your day.
Or was that all on Thursday morning?
I absolutely guarantee it will change your spending habits this coming decade.
I'm not sure if there's a way to understand until you experience it being walking distance from your office.
I suppose it is not so likely there will be a store in walking distance the next few years.
Shopping at the Go Store is much faster. When you're grabbing a single item you can have times of ~10 seconds. Grabbing more items and you save more time by not having to check out. No checkout, no lines, no figuring out why your phone or card isn't accepted, or what error message is on screen etc. In and out as simply and efficiently as possible.
I wouldn't say it is revolutionary, but it does cut down the time it takes to go shopping. It's like a perfect version of self-checkout.
Another, more subtle feature is that I never wait in line because I have trained myself to never go at peak times. I waste a certain amount of time thinking about it.
That said, I find the idea of a cashierless store super-creepy and will be among the last adopters.
Using vision to solve a problem always seems inefficient but it scales out way better than anything else (ie the various ways google maps uses vision when in a world with perfect map data it wouldn’t be needed).
"Unidentified item in the bagging area!"
"Please place item in the bagging area!"
How would you like to pay? Pick credit card Please complete purchase on separate device to the right.
Do people really love even those voice / flow of these machines?
I've used amazon go. It is as far as I can tell very low friction. I'm not 100% sure how they figured out what I got but it was correct in my case.
Also, the Scan and Go system they have at Sam’s Club is awesome. I can scan items on my phone as I’m shopping. When I’m done, I tap purchase, and then walk out with my stuff.
To be allowed into the store requires a few prerequisites which unfortunately not everyone has the benefit of possessing; smart phone with data service, credit card, amazon account.
For most people this is no problem. But what about all the people who rely on cash based systems to exist? What about people who don't have smart phones with internet plans (though this is becoming an increasingly small population, it's still very prevalent)?
Now what happens to those people when this becomes a standard and the alternative options diminish? Food deserts in underserved (poor) neighborhoods are already wide spread in the US, this credit based system which relies on an app will only make it harder for people who already suffer.
The one problem Amazon would have here is preventing customers from leaving if their purchase exceeds their cash balance on the card. Seems relatively easy to stop with a 2-step system for exiting (cross a line/barrier to signal that you're leaving, wait for a physical barrier to open after the payment method is debited).
That being said, we have plenty of grocery stores in my area and they seem to compete quite well with Costco. I am not worried about these stores, and I don't plan on using them because I find them to be a privacy issue. I think there are plenty of people who either prefer cash or don't want to tie their food purchases to their Amazon account to keep traditional grocery stores in business.
Then I go to a place like safeway and literally everyone pays with a credit card and gives up their phone number so they can get MAJOR discounts on purchases (ie, $1/item vs $3/item) and points towards things. Even the folks on EBT are using some kind of credit cards instead of the old vouchers.
I believe actual stamps have been gone for over 20 years at this point.
But people are screaming for just the opposite with stores that don’t accept cash. They want government intervention.
These stores are now < 0.1% of all stores. Worrying about what might go wrong if/when they're 90% seems very premature.
Just look at Amazon.com vs Walmart, Walmart is still in business, even though running a brick and mortar store is more expensive than an online store.
Part of this is security - walking around with lots of cash / keeping it at home (which may be a shared room) is just not workable.
Costco, Sam's Club, and other similar 'big box' stores are usually located in less-dense areas and are not competing directly with 7-11s or other convenience stores.
It’s not hard to get a cheap smartphone (something like 60-80% of homeless people own a smartphone based on a quick google) and if you qualify for SNAP, you can almost certainly get a free smartphone through the lifeline program.
> data service
Amazon Go stores have free WiFi.
> credit card
Edit: Originally stated you can use gift card credit, but that’s incorrect. You can get a secured card with no credit check though.
> amazon account
Free and obviously easy to get if you have access to the internet, which virtually everyone does through a public library.
https://www.pewresearch.org/global/2019/02/05/smartphone-own...
75% of Americans have credit cards.
https://www.creditcards.com/credit-card-news/ownership-stati...
10% of Americans use only cash.
https://www.creditcards.com/credit-card-news/payment-method-...
62.7 million Americans don’t have smart phones
82.5 million Americans don’t have credit cards
33 million Americans use only cash
Another way to look at it
The equivalent of the entire population of the United Kingdom not having smart phones
The entire population of Germany not having credit cards
Nearly the entire population of Canada using only cash
The Amazon Go convenience store at 920 5th Avenue (looking from the outside) felt a bit sad in their dining setup - it's a tiny bar that seemed designed for quick, alone consumption. I'm guessing most folks going there grab a lunch then go back to their office to eat. (image: https://www.google.com/maps/place/Amazon+Go/@47.6063189,-122... )
Is this to the exclusion of debit cards?
In the UK we have at least three tiers of supermarket.
You're not going to see the riff raff in Waitrose because they can't afford it. "Above" that you have more esoteric health shops.
You might see middle-class folk who are cost-conscious (waves) in the low end supermarkets.
Pound shops and discount retailers tend to have a specific style - music, decor etc - that reinforce the fact that they're 'poor people stores'.
Sure, theoretically you could go and buy your groceries in the expensive stores if you're skint, but in practice no-one actually does that.
The store itself was nice enough. But it was small, and felt like a novelty, like a lab experiment and not an actual convenience store. As a software engineer, I was also keenly aware of the massive array of tracking technology all around me, and that didn't feel great.
I don't really have any complaints. It was convenient, but not "oh wow this changes my life" convenient. If this kind of technology pushes down the prices (though that's honestly hard for me to see how), that will be a benefit.
The one in the West Loop is right next to the trains and sees a lot of traffic during rush hour. So they have 4+ employees just standing around watching people. The ones at random locations in the loop aren't as busy - so only have a couple employees in those stores.
Actually I think it's weird that Amazon has so many employees in Go stores. You'd think they'd just need 1 person to stock things and call the police if someone jumps over the gate without scanning their phone. Every time I see a group of their employees standing around I want to ask "why haven't any of you been replaced by robots yet?"