A friend of mine works for a large battery company. The moment the hurricane forecast shows one headed for a populated area they start packing and dispatching semis of AA/C/D/9V batteries to all the Walmarts, Targets, Grocery Stores, etc in that area. I'm sure the bottled water and plywood manufacturers do the same. How does this differ?
Those are examples of predicting demand changes on a macro scale.
Amazon wants to do this on a more micro, person-by-person level. For example, you visit a product page for some specialty shampoo. Amazon might decide to ship that product before you even add it to your cart.
My response to Amazon, if they do this, is something I call predictive returns. As in, I predict I will return everything they ship to me in this manner.
I will go on record as saying that this is similar to Amazon announcing drones. It's part PR move but true there is an obvious connection to reality.
Hard to believe that Amazon is going to move the needle much with something like this. Shipping is already fast. Getting it where it needs to be is obvious. As you have noted there are many examples of manufacturers anticipating demand and doing something similar. But they don't have or file for patents. What about a patent for anticipating that your chain restaurant will need less help on certain days of the year or when it rains? Most business would not waste time on this thing (patents) although they do these things if they can. They don't crow about it though (unless they are doing a dog and pony show).
This is the type of thing that most companies would do in many cases and never talk about. Because they aren't as obsessed with creating an image of invincibility which Amazon spends so much time with. [1] It's kind of like the shock and awe of US Military propaganda (where they show the well oiled military machine.) People eat this type of thing up. "Oh look how special Amazon is".
[1] To scare off competition. Look how mighty we are. Look at the barriers to competing with us!
I don't think they are improving shipping speed to compete with online stores, as much as to compete with stores in your community. I think the goal is to make ordering from Amazon almost as quick as getting in your car and going to a local store.
The benefit I see with this is they can snail mail it to the local hub so total shipping costs would be lower for Amazon when you do actually order the product even if you're using Prime.
Because there's no 'National wants a 24" Asus 1ms thin-framed LED gaming monitor, Logitech gaming controller, 5.1 surround audio, and ATS Acoustic Foam Corner Base Traps Forecasting Service'. It's logical to pick out geographic major events from current services and ship batteries, tarpaulins, baby formula and bottled water to them. But for everything else, you need fancier machine learning and lots of training data, and not everyone has that. Walmart doesn't have my wish lists. Amazon does.
Edit: Before anyone mentions it, don't buy those bass traps. They're not big enough to do the job for low frequencies. For that you need a pallet of rockwool and a weekend carpentry project ;-)
I agree with you - seems very similar. They may use data from individuals, but they will aggregate it - needs to be a relevant sample size with some kind of pattern to it. IMHO, having a flat panel TV in my wishlist is probably not reason enough for Amazon to move it to a local warehouse. The biggest category of things that still get my local dollars are stuff I buy at the drugstore, that I sometimes buy bulk on Amazon when I think ahead. If I knew I could get it same-day from Amazon on Saturday ... I'd continue to sit here reading hacker news instead of putting on some pants.
This type of plan seems a lot more reasonable when you picture it coordinated with a marketing plan or advertising campaign. Given Amazon's analytics, they can certainly target a certain geographic area with an ad and predict the response rate of that campaign, while the estimated quantities of product are already staging at the local warehouse.
Haha funny, but this is so very different - I can't believe how many people on hacker news can't wrap their heads around the distinction between this 'anticipatory shipping' and retail.
Amazon specializes in providing the long tail (http://en.wikipedia.org/wiki/Long_tail) to customers; Walmart is the utter opposite. By aggregating predictive data about the cumulative micro-consumption habits of users, shipping times could be lowered for the products inside the long tail.
If you don't want to compare to Walmart, compare to a mom-and-pop store who stock products that they know local customers will buy, sometimes down to an individual customer level.
This is traditional stock management, pure and simple.
"To minimize [the cost of returns], Amazon said it might consider giving customers discounts" -- just like any other overstocked store.
Absolutely incorrect. The long tail of products and the size of a store has nothing to do with each other. A mom and pop store is not going to cater to a local population any better just because it is small. It's hilariously wrong to think that people's buying habits are localized or consistent based on geography.
If I need a linear carriage, a Razer Hydra, an obscure philosophy text by Merleau-Ponty, a 1/8" straight fluted HSS router bit, a jar of Mexican mole sauce, and a monitor arm mount -- where do I go? Walmart? My neighborhood mom-and-pop store? Of course not.
The difference is that this is a way to get stuff to your door within a few hours of your placing an order, instead of you going somewhere very 2 weeks.
Most local shops around here don’t mind delivering the item you want to your home after a phone call, though likely at a heftier premium than ‘standard’ online merchants.
Do Walmart items ship out of your local superstore or do they ship from the regional hub? Assuming you buy from Walmart.com and don't wander the miles of aisles in the store.
You've always had to pay tax on online orders, assuming your state has a sales tax. But no one ever bothered to file the voluntary form, costing the states lots of money that they should have by law.
So states such as Virginia have cut deals with Amazon to collect tax from them. Congress is talking about making it easier for other states to do the same. The era of the tax-free internet is nearing an end.
This is a patent for turning UPS and Fedex distribution centers into warehouses. They're preloading the distribution system so that the only latency is the last tens of miles, rather than hundreds or thousands. It could be same day assuming you got your order in before 4am or whenever the trucks roll. Or even later if they have a more courrier oriented partner.
Dell currently does something very similar. If you have "same day 4-hour enterprise support", then the replacement part will most likely come from a distribution center collocated with your local courier company.
Ah, similar to Amazon's "shopping cart," the shop could provide a physical cart (imagine a cross between a baby stroller and a faraday cage) to help customers carry inventory to the exit of the store.
And if the customer is too busy to visit the physical store, or just doesn't feel like it, Amazon could put the pictures of all of the products they have on a web site, and simply mail the customers their selections.
That's a pretty good idea, but don't you think it would take too long for the shipments to get there?
They should come up with a way of tracking what pictures people look at and then shipping the stuff in the pictures to the warehouses near those people. I bet you could do the picture tracking thing with an ActiveX control. That wouldn't work in Navigator, of course.
We could get around that issue by having a dedicated program for looking at the pictures. We will develop a networked software delivery system that people will use to install this program quickly and easily.
I see where you're going with that, but I don't think delivery of software over a so-called "network" will ever be feasible. Even with a 1200 baud modem, transferring a multi-kilobyte .exe would take hours, and that's assuming the BBS doesn't drop the connection partway through!
It seems more realistic to transmit the pictures using CD-ROM technology: it's like Netflix, but for CD-ROMs. Amazon would mail you their catalog CD-ROM, and you look at a slideshow of all the products they sell. Then, when you decide what you want, simply fax your order to Amazon, and they tell you which "shop" to drive to in order to pickup your merchandise.
Why stop there? Similar to amazon one click purchase, if customer has amazon credit card with rfid chip, they can simply pick up the item and walk out of the store - it would be charged auto magically.
Unfortunately, reasonable customer-oriented space is much more expensive than a (semi-robotic) dimly lit warehouse somewhere on the industrial outskirts of a city.
Uhm, operations research is what, at least 75 years old now? And this is exactly classical operations research stuff, at least the idea is nothing new:
It's almost like Amazon is recreating the local retail experience, except that all the mom and pops are owned a monolithic entity with no ties to your community and you can't actually check out the products before you buy.
I'm heavily opposed to winner-take-all and so my boycott of Amazon books must continue. I don't really buy anything other than physical books off of Amazon, so it's not a huge sacrifice.
What if everything you bought from Amazon was automatically available for resale used after a certain period of time? You're thinking of buying the $79 textbook, but Bob who lives across the street will sell it to you for just $39 (Amazon gets $10 for facilitating the transaction, but will take the risk) and Amazon knows he's home right now.
Or you borrow it from him for a week for $19, he'll also lend you his blender for another $19.
My brain really wants this to happen, but my wallet doesn't. That would be really effective! If they really could get the costs right maybe it could happen - of course, the math just might not work out.
Isn't it a kind of self-fulfilling prophecy? As in, Amazon make a loose prediction about what you'll want to buy, but the hook of having that one available within 1 hour biases you to be more likely to buy that item. Then Amazon get to say "we're amazing at prediction!" whereas they're actually amazing at altering customer behaviour. Which isn't to say that that's a bad thing, it's just a bias which needs awareness.
I imagine transportation costs could quickly spiral out of control if you didn't have the analytics tuned correctly. But one particular use case comes to mind: you look at a couple items, maybe even add them to you cart, then walk away. A lot of retailers now will send you an email for "abandoned carts"—imagine if that email also included a line that said, "Oh, and if you finish your order, we'll have it on your doorstep tomorrow." Pretty compelling.
Indeed, there is a lot of variable to consider and tune, each with it's own pros and cons. Unfortunately, whichever of the many variables you consider, and however you tune them, you'll be infringing Amazons patent.
They got a patent for this? "People in this city order this book more often than other cities. Okay, let's make a deal with our shipper there to stock some onsite."
I'm not sure if you're being deliberately hyperbolic, but I think your example is an extreme oversimplification.
More like "Joe Bob in Nowhereville, KY has this item in his wishlist. He looked at it three times in the past three days, and twice today. The last time he exhibited this pattern he bought it within two days, as have 78% of the people in this geographical region who have exhibited similar patterns. Let's ship to a local transit in prep for his order."
Actually really cool - and at least based on the summary (can't view the linked doc b/c of missing plugin of some kind) it's a business method patent as opposed to a software patent.
"anticipatory shipping", this is pretty much just saying that local warehouses/stores should anticipate demand by using as much data as possible. For Amazon that means wishlists, viewing history, shopping carts, emails etc. This is good Jeff Bezos, but not good enough. I could do this better.
In order to best anticipate demand, they should be trying to make deals with manufactures instead. This system would initially work best with subscription based products. And then we could figure out how to turn everything into a type of subscription.
Manufactures have a release cycle that is important to them staying 'competitive'. So they have to constantly make new models and 'retire' the old ones by planned obsolescence forcing us to buy the new slightly improved model that has a time bomb built in it to go off when the next model comes out.
So basically Amazon needs to place their customers in a 'release preference' category. With names like, 'bleeding-edge','beta','stable','2 generations back', or 'annually/biannually'. This could make manufactures more responsible and get the latest designs out of secret labs and into consumers hands quicker.
Customers can choose which category they want to be included in for different product families. So for example, I might choose the 'bleeding-edge' release cycle for my phone, but prefer a biannual subscription for a refrigerator or car.
And unless someone has a brand preference, the best products can be automatically chosen based on reviews from amazon and social networks and blogs and benchmarks on hardware/battery life/speed/etc
Discounts will be given for recycling your previous models, and as our 3d printing becomes better and closer to home, eventually we might just get 'ink' credits to print out phones, computers, cars, furniture, TV's etc.
On a side note, this kind of system would be the system of all systems and it would be hard for competitors to start (without the data owned by Amazon). So I don't think 'Amazon' should get to boast and claim this system as their own and keep the profits. I would like to see this type of system organized at the national level at first (until we fully embrace globalization) and we could vote on an open source 'shopping and distribution' system whose profits go back directly to the people in the countries they operate. Then people could join and accept a national system to be proud of, one whose contributors, ideas,transactions,money,etc are all open and viewable by anyone to encourage accountability.
>> The patent exemplifies a growing trend among technology and consumer firms to anticipate consumers’ needs, even before consumers do.
Wrong. The patent exemplifies a growing trend among technology and consumer firms to think their business processes ought to be protected by patents when in fact, no such need exists. Certainly there is no wider (public, society, industry) good from allowing rich companies to patent something many others may have easily figured out independently.
Do you have any evidence that Amazon intends to use this patent purely for defensive reasons only? If not, we're just speculating. Worse, we're giving them the benefit of doubt.
A method and system for anticipatory package shipping are disclosed. According to one embodiment, a method may include packaging one or more items as a package for eventual shipment to a delivery address, selecting a destination geographical area to which to ship the package, shipping the package to the destination geographical area without completely specifying the delivery address at time of shipment, and while the package is in transit, completely specifying the delivery address for the package.
As long as you consider being at the store "in transit" they've effectively "invented" physical stores with a digital only interface. The patent even describes a method of giving customers in the region discounts on products that don't sell and are sitting in the "store".
There is nothing novel or new about this and the patent system is broken for more than just software. The entire rest of the patent is a bunch of filler to make it seem like there's something important going on when there in fact is not.
Building warehouses closer to customers isn't that far off "anticipatory shipping." Could "anticipatory shipping" be a clever scheme to reduce shipping times while also avoiding establishing tax nexuses (and therefore avoiding the requirement to collect sales taxes) in additional states?
Yeah, but that isn't exactly news. Or patentable. (The details of having the analytics fine-tune how much of what to push to which edge node may be patentable under current law, though...)
Perhaps I'm wrong, but I thought Netflix already did this with their DVDs. They would see that certain DVDs were near the top of the queue in certain areas and move the DVDs closer. Works especially well with more rare items.
It may pressure consumers to decide to buy now, if they get visibility on a desirable product approaching closer to them in delivery time. Especially if they are also aware that the approaching product may be snapped up by a rival local consumer with similar tastes. Game theory applies. And this will also be gamed by customers who understand Amazon's cost and incentive economics.
Smart move though, as it can focus customer awareness into their distribution chain which is locked-in. Attention shifts from logical: ProductX() to instantiated: new ProductX(Amazon_supply_chain).
This will work great for things that people "subscribe" to like toilet paper and dog food, which, I kid you not, is cheaper via Amazon than locally, and, someone else lugs it to my house for me.
"So Amazon says it may box and ship products it expects customers in a specific area will want – based on previous orders and other factors — but haven’t yet ordered. According to the patent, the packages could wait at the shippers’ hubs or on trucks until an order arrives."
Here's an idea. Pre-ship merchandise to local independent retail locations, waive the shipping fee but let the retailer mark it up a bit, and offer in-store pickup.
It's a proven business model, since that's how people bought stuff before catalogs and the internet.
This is not shipping before you buy. This is online interface for a brick-and-mortar store (such that many overheads of traditional stores are gone). Call it a distributed warehouse if you like.
Patents Office got fooled into seeing this as an invention. The violation could happen only if they could predict what I specifically want to buy and ship it for me specifically. If all they do is optimise distribution, well that's happening since the dawn of trade may I guess.
So does that mean someday I will be able to call in a same-day airdrop of semi-uncommon product x that has been sitting in my amazon shopping cart or wishlist?
An interesting tactic. If you know you're going to want something but don't want to pay for it just yet, add it to your cart a week or two in advance to let it "preload". Real-world precaching.
'“It appears Amazon is taking advantage of their copious data,” said Sucharita Mulpuru, a Forrester Research [FORR -2.03%] analyst. “Based on all the things they know about their customers they could predict demand based on a variety of factors.”'
Something about seeing a big red box next to the analyst highlighting their company's negative stock performance makes me a wee bit less confident in their analysis.
This definitely makes sense for their subscribe and save items. If they're not doing it already for them. But I don't think they are. I've got quite a few things on subscribe and save, I get charged the last day of the month and it still takes about a week to get my items.
This makes sense, especially if combined with a local retail channel. It would be neat to be able to travel somewhere and get hooked on some obscure local food/drink and come back home to find that you're local shop now has it in stock automatically.
So Amazon would basically fill an actual shopping cart with your current order at a hub close to you, and when you're done paying, would just come out of the neighbour's driveway and wheel the cart to your door...
I actually find this clever and patentable. It's machine learning assisted caching for physical packages. Now it all comes down to the predictive performance of their algorithms.
Amazon once sent me the wrong parcel by mistake. I expected a long convoluted returns process, but instead Amazon just said I could keep all the items and that anything I didn't want, I should give to friends or a local charity.
It's probably cheaper & more time-efficient for Amazon to write off the cost of the item than to deal with reshipping & restocking an item that can't be sold as new anymore.
