EP054 - Amazon Q3 Results, Money2020, IBM World of Watson
EP054 - Amazon Q3 Results, Money2020, IBM World of Watson
Episode 54 is a news update:
Amazon
- Amazon Q3 Earnings Call
- Rumor Amazon could open 2000 grocery stores
- Amazon capturing lower income shoppers
- Amazon Fresh expanded to Chicago and Dallas
- Cowen and Company: 50% of US Households could have Amazon Prime
Trade Shows
Other News
- Customer expectations for fast shipping getting faster
- Piper Jaffray Millennial study
- 2016 Deloitte Holiday study
Episode 54 of the Jason & Scot show was recorded on Thursday October 27, 2016.
Join your hosts Jason "Retailgeek" Goldberg, SVP Commerce & Content at Razorfish, and Scot Wingo, Founder and Executive Chairman of Channel Advisor as they discuss the latest news and trends in the world of e-commerce and digital shopper marketing.
0:02 Welcome to The Jason and Scot Show, your source for the latest news and trends in the e-commerce industry. Featuring host Jason "Retailgeek" Goldberg, SVP of Commerce at Razorfish, and Scot Wingo, founder and executive chairman of ChannelAdvisor. Here are Jason and Scot Welcome to the Jason and Scot Show. This is episode 54, being recorded on Thursday, October 27th, 2016. I'm your host, Jason "Retail Geek" Goldberg, and as usual, I'm here with your co-host, Scot Wingo Hey, Jason, and hey, Jason and Scot listeners.
0:43 Welcome to the Jason and Scot Show. Jason, how was Las Vegas? Las Vegas was awesome. Uh, although I was thinking of you because I know there was an Amazon earnings call going on while I was there, and I could just imagine how excited you must be to do a podcast. I was. I was, uh, on the seat-- on the edge of my seat on-- wanting to tell folks about what was going on. I also definitely wanna hear, you were out there for Money 2020 and an IBM show. So I wanna hear, uh, definitely wanna get a scouting trip report from you on that. Uh, let's jump into the Amazon news. So, uh, as we record this, it's the same day Amazon released their third quarter earnings, and that's always an important one 'cause it sets up, uh, what the rest of the year's gonna look like as far as the holiday.
1:28 Um, so I felt like it was a pretty strong quarter. Amazon's growing twenty-nine percent. I always use the benchmark of fifteen percent, which is comScore's number for e-commerce growth that includes mobile and desktop. So that's easily two times as much. It was a little tick down from last quarter. So in Q2, Amazon grew, let me make sure I get this right, thirty-one percent. So this was twenty-nine percent. So you may see some headlines, you know, growth slows at Amazon or something like that. When you peel the onion on that, though, Amazon has two categories they report on: media and EGM, which listeners of our Amazon deep dive will know stands for, uh, electronics and general merchandise. So effectively, that's the part of, uh, un-unless you're listening to us from, let's see, Barnes & Noble, that's the part of Amazon that everyone really focuses on.
2:15 Um, so, uh, the-- that actually did not slow down. It was pretty consistent. It was thirty-four percent in aggregate, uh, and, you know, another really strong quarter there for EGM. Um, third-party sellers did really well. They hit a new high watermark here. Fifty percent of the units, uh, uh, GMV, I haven't done the math yet, but I'm, I'm thinking it's gonna tick over sixty percent as a GMV number, which is pretty interesting, uh, are now third-party sellers. Uh, and unit volume overall was twenty-nine percent growth. So, so I, I felt like it was a pretty strong quarter. Uh, it does include Prime Day, uh, in the quarter, uh, and, uh, I think the setup was really good. Another thing you'll probably see in the headlines is Amazon gives this very broad guidance on the bottom line or profitability, and they came in within that guidance, but Wall Street had kind of thought they would beat that guidance pretty handily because they have in the last couple quarters.
3:09 Um, so Amazon has kinda had this really nice progression of profitability. Uh, Wall Street was getting really excited about that, and they kinda had, had projected that third data point to be up and to the right, and it kinda went down a bit. And, um, you know, the-- on the call, that's where most of the questions centered, and Amazon had a pretty good answer. They said: Look, we, uh, we are We've announced, uh, and have built 18 fulfillment centers, and we've got another seven coming this year. So that's 25 fulfillment centers this year, up from 12 last year. Uh, and then they, they gave a new data point. They said that this will represent an increase of 30%, uh, in square footage, uh, that they're building this year, which is, which is pretty w- when you think about the scope of what Amazon has, it's, it's pretty amazing to be building that much fulfillment center.
3:57 Um, they're already got like, you know, eight to nine times more than Walmart or anyone else, and this will get them up kind of, you know, very much north of that. Um, they also said there's a lot of other investment areas right now. They've got India. Uh, you'll like this. They called out Echo and Alexa as an area of, of continued large investment in AWS. Uh, they also have doubled the amount of video content that they're acquiring and, and developing. So, um, you know, a lot of people thought that this whole, uh, Amazon Prime Video was an experiment, and, you know, gosh, can they compete with Netflix? How serious are they? And they seem to be quite serious. These numbers are, you know, astronomical, uh, in the hundreds of millions of dollars that they're spending on video and things like that.
4:38 Uh, so all in all, it was a, a really good quarter, I felt, and it really teed up pretty strongly for fourth quarter. Their guidance for fourth quarter was, um, I'll have to pull it up, but it was kind of in the, in the high teens, so kind of like 19% year-over-year growth. I, I, I feel like that's being relatively conservative, and, and I think they should beat that pretty easily, but we'll have to see how that comes out. One thing you'll like, they, um, because they're, um… you know, we've talked about it on the podcast a little bit about this project X or what's that other name that we've heard? It starts with a C. Is it Como? Como. Um, a lot of the analysts on the call were like, "Well, is Amazon Fresh dead?" And, uh, they said, "Oh, no, no.
5:19 Fresh is still kind of a core offering for us." Um, my, my reading the tea leaves is they kind of view it as working great in large metros, and some of the other things they're experiment- experimenting with are, are kind of more, um, suburban. Uh, like Prime Now is available in a much more sub- suburban footprint. And then they did announce, you know, this will make you very excited, uh, that Amazon Fresh is launching in Chicago and Dallas. Uh, I don't know if it's immediately available, but they said it's, you know, imminently available. So you should have to t- you should check that out, and maybe on the podcast we can get a report of your impressions of Amazon Fresh there in Chicago. I will. I, I hope to make a delicious Amazon Fresh meal that I eat during a podcast. Oh, that would be, um… I'm sure our listeners would enjoy that.
6:02 How could they not? Um, and then you h- were telling me, um, uh, about some other grocery news you heard about. Yeah. Well, the, so there's a couple articles out there. They're not perfectly sourced at the moment, but there are a couple articles talking about some internal documents at Amazon that reference, uh, plans to have, what, 2,000 of these grocery pickup locations by 2020 Um, and so I think that raised a lot of folks' eyebrows that, that, you know, that's a, a footprint, you know, very similar to a Kroger or a, um, you know, a Walmart or, or w- or one of the other, you know, folks that they're likely to be directly competing against in the grocery space.
6:43 And it, I-- to me, if that's true, that just reinforces what, what we've said in the last couple shows about grocery and that, like, the, the large volume, uh, use case for grocery is not gonna be home delivery. It's gonna be this order online pickup, uh, at, at a venue and, and that they recognize they have to have venues that are more convenient to more customers to, to really drive it. But at the moment, like I don't think there's any evidence that Amazon's out there signing 2,000 leases. I think, I think we're gonna see a, as we often do with Amazon, a controlled experiment and, uh, if that, if those experiments go well, like will Amazon scale it to, to national, um, scale?
7:24 Like of, of course they will, and I, I don't think that would surprise any of us. Yeah. The one article I read kind of s- talked about a 10 to 20 store pilot program and, and in there it talked about a couple different formats. One of them is that Project X, which is a drive-through pickup kind of thing. Uh, and then it also talked about having a traditional kind of, you know, um, rolling a cart through aisles. They called it a destination concept. Uh, it's kind of a weird name for it, so I, I feel like maybe it's a little bit more than your average Kroger or something. But, but it, it does seem like they're gonna be testing a bunch of different formats for, for figuring out grocery. Yeah. I, I, I'm gonna be really interested to see what they experiment with and see if they sort of invent some new shopping experience. That would be really exciting.
8:05 Uh, you know, another thing that, that struck me as I saw all of the, the, this earnings data, and, you know, Am- Amazon's just getting bigger and eating up a, a bigger chunk of American consumers' wallets. Um, there was another report out, and, uh, I'm, I'm ashamed to say I can't remember the source right now, that was showing the average income of Amazon shoppers going down. Um, and that's because they're, they're just capturing more consumers at, at lower income levels. Um, and that, you know, that's an interesting thing. You know, for a long time there were these talking points that, like, the Walmart and Amazon shopper really didn't overlap because Walmart had these lower income shoppers in the middle of the country, and Amazon had the affluent shoppers on the two coasts.
8:47 Uh, but it, it seems like these guys are clearly on a collision course, both with their overlapping efforts in grocery and, you know, that they're, they're definitely starting to overlap the customer bases more and more. Yeah, that's, uh, John Blackledge over at Cowen, and he does a, an annual survey of, of internet consumers, and he gets a lot of good Prime data there. And, uh, they're definitely… You know, I think what's happened is, and Gene, Gene Munster, uh, over at Piper Jaffray has a similar kind of thing he does. He's already done his this year. Um, and, you know, they-- What, what they're showing is they've really kind of… The US Census Bureau takes families and, and gives them a household income, and I think the highest bucket's one, 125K.
9:28 Then there's, like, an 85K, a 65, a 45, and, and below. Uh, and, you know, what you see is Amazon's gotten a really, really high concentration at that top bucket, but now they're really kind of working their way down. And as they do that, um, that, that's starting to lower the, the ticket, the lifetime value of those customers as they go further down that spectrum. Um, and a lot of people speculate that's why Amazon did, you know, breaking, uh, allowing you to buy Prime in kind of a monthly increment. So the, the hundred dollar kind of upfront nugget was really kind of oriented towards those higher income families, and then now they're working their way down. And, um, it'll be interesting to see if, if the, you know, their audience does collide with, with the Walmart audience.
10:11 Yeah, absolutely. And, uh, I feel like there were, and maybe it was from, partly from that same report, but, uh, there were also a couple news tidbits this week, uh, with new Amazon Prime estimates Yeah, that one's from, uh, it's ca- CIRP, CIRP, and it's like the Center for Individual Research something. Um, they, they came out at 65 million, which, which for the US, which feels high. I think, um, most of the folks-- So, so our estimate is 60 globally, which I think gets us to 40 to 45 in the US. So I've always felt like that was a little bit high. Um, but you know, uh, the, the headline from that one is 20% of American households are on Prime, uh, which, you know, I, I think it's probably closer to 15 or 16.
10:59 But, um, it's still, you know, a pretty material amount and definitely a very popular program. Yeah. Uh, there was also, you know, the Piper Jaffray survey of millennials this, uh, that came out this week or, or a couple weeks ago. And, uh, one of the data points in that was that like 50% of the respondents in that, so 5,000 of the 10,000 millennials were in Prime households. Hmm. Yeah. And then, um, that's a really good report if, if, uh, folks that are listening are interested in millennials. Another thing they do is, um, he kind of ranks various brands, so, so like apparel brands and other things like that. And, uh, also he has a, a question in there around retailers, you know, which retailers would you consider shopping at or have you shopped at?
11:45 And, um, you know, it's kind of some of the usual sus-suspects at the high end, but I, I think one of the-- I don't know if you saw it, but one of the funny ones is out of something like 10,000 millennials, only, uh, you know, one or two selected some of the other retailers that are pretty popular. So, um, you know, definitely very different demographic in, in how they behave and think about these brands that we've, we've grown up with. Yeah. I mean, I mean, I think Walmart was one out of the 10,000 and Target was like six out of the 10,000. So that's, you know, pretty substantial shift in, uh, consumer behavior. Yeah, and we'll, um, we'll, we'll put a link in the show notes to both of those reports in case anyone wants to, to dive into them.
12:26 Uh, yeah. And, uh, in a perfect tie-in, uh, we- very soon we'll be releasing an interview with, uh, the president of e-commerce from Steve Madden, and the, the millennials named them as the, the fifth… their fifth most favorite footwear brand. Yeah, yeah. So enough Amazon news. Tell us, give us a little trip report. So you were at Money 20/20, and, uh, the IBM shows, uh, that's their new World of Watson, right? Exactly. I keep calling it World of Warcraft, and they keep correcting me. Uh, but we'll- You should have worn your, uh, mask there. Exactly, yeah. The b- They're, they're very grateful that I didn't come in costume. Um, but so as usual, the, the e-commerce industry is extremely nice to me, and so they've, they conveniently now are co-locating shows so I can travel to Las Vegas for a week and attend two shows instead of just one, which I, I certainly appreciate, although I was doing a lot of back and forth.
13:20 Um, but the Money 20/20 show was at The Venetian this, this year. Uh, that's a show we've talked about in the past. It's primarily about, um, payment technologies and, and, uh, financial services tech. Uh, I think this might have been the fourth year, and the show's grown very rapidly. There were probably about 12,000 attendees this year. Um, they, they started that show at the Aria Hotel in Las Vegas, and last year they outgrew the Aria and had to move to The Venetian. Um, so big show. Uh, in years past, there was some like, you know, controversial new payment technology that was launching. And so, you know, two years ago there was a ton of talk about, uh, Bitcoin, for example. And, and last year there was a lot of Apple Pay talk.
14:03 Uh, it didn't feel like there's some net new exciting technology this year. You know, it's the, the, the new technologies have matured a little bit. So it's a lot more talk about EMV adoption, which, you know, EMV is the Euro MasterCard Visa standard with the, the chip and signature here in the US that's- Somewhat of a mess, and, uh, I'll, I'll put a link in our show notes, but, uh, I have a Instagram feed where I just upload pictures of the signs that people have had to tape to their cash registers to explain to customers how to, to check out because we're so, so convoluted in the US at the, at the moment. So that's kind of fun for me.
14:44 But there was a lot of talk at, at Money 20/20 about resolving that. If any of you have, uh, ever, uh, checked out in the US with a, the chip and signature, one of the things you'll note is it's much slower transaction than the old credit card swipe used to be. And so a lot of the, the payment gateways are now working on this new technology where you can, um, dip the card much faster, so you don't have to leave the card in the, the terminal for the whole transaction. You just dip it, it reads it, and then it does the transaction, and the transaction actually takes the same amount of time, but consumers perceive it to be much faster because they're putting away their credit card and doing something else while, while it's processing.
15:25 Um, so that's interesting. The, uh, if, if you walk around this show, you're gonna see a ton of these fancy new smart, um, payment terminals that, that everyone expects to replace the old POS terminals at a lot of retail stores. And because retailers wanna be compliant with EMV, that's, that's driving retailers to invest in new hardware, so there's this big, uh, inflection point in the POS industry, and they're all selling this new hardware. Um, and a ton of this hardware is based on the Apple architecture, which is pretty interesting. So, you know, a lot of these things are, are based on consumer technologies where they're expecting you to snap a iPhone or an iPad into a, a device as opposed to getting a dedicated terminal from a, um, you know, a Verifone or a Poynt or, or one of these, these big smart terminal manufacturers.
16:15 Interesting. Did you see, um, also today, it was a big day. The, uh, uh, Apple had their keynote, and they released a whole new family of Macs, uh, and they updated the Apple TV. Uh, and it was interesting. The, the rumors were true that, and I think we've covered it on the show, they, they are gonna have Touch ID on the Macs. It's up by this new thing called a Touch Bar. Um, and they, uh, they demonstrated… I forget what they bought. Um, they demonstrated someone purchasing something live on stage off the web using Touch ID, which was pretty cool. Yeah. Now, I'm gonna ask our listeners, don't be that guy that brings your laptop to the re- the retail store and pays with Apple Pay at the cash register using your laptop Yeah, use your phone, folks. Come on.
16:55 Yeah. Phone or watch Use the smallest, smallest screen possible. Um, if you go to like the Visa booth, uh, for example, they, they're showing a whole line of jewelry that you can pay with that have sort of em- embedded NFC technology in it. So they're, they're envisioning this world where you have a NFC ring or, or necklace or, or a bracelet or something that you can, you can pay with. Cool. Did you speak at Money 2020 or you were mostly, uh, mostly in the audience? I was an attendee at Money 2020. Uh, I, I did speak at the IBM show, which we'll talk about in one minute. Um, the biggest announcement that I think is relevant to our listeners at Money 2020 was that PayPal is now, uh, available and embedded in Facebook Messenger So, uh, we've, we've talked a lot about chatbots for customer service and chatbots for commerce, and how they've been very successful in Asia, and they, they haven't seen great adoption in the US.
17:53 And, you know, generally, my hypothesis, and then also some smart people's hypothesis, has been that the US chatbots were lacking sort of two, two pieces of infrastructure that the Asian chatbots had, right? Like, one was APIs to enable third parties to develop their own, their own bots, which, uh, Facebook in particular launched about six months ago. And so they've gone from zero third-party bots to 18,000 bots on the platform now. And then the other thing was, uh, a broadly implemented mobile wallet with a bunch of consumers' payment information already stored in it. And, and so, you know, that now comes to Facebook via, uh, PayPal. And so it'll be really interesting to see, you know, there… When, when PayPal first launched the API, there were a bunch of retailers that launched these, like, early commerce experiences, and I think, like, KLM Airlines and 1-800-Flowers are two that come to mind.
18:45 And they were still kind of clunky experiences that you wouldn't necessarily prefer over some of the old ways of shopping. Um, but now that we have, uh, low-friction payments built into the platform, it's gonna be really interesting to me to see if they can, uh, uh, start to, to mimic some of the success we see from WeChat and some of the other platforms in Asia. Yeah, it seems like it would be the use case of sending money to a friend, like a Venmo kind of thing should be very easy, right? So you could just message and say, "Hey, I want to send Jason 10 bucks to pay him back for whatever." Um, but I think integrating PayPal into the bots is gonna be a little trickier and, and almost a little bit at conflict because I think Facebook also has their own transactional piece there.
19:30 So that's gonna be interesting to watch to see what that looks like. And do you end up with that NASCAR kind of thing you always talk about, where I'm, I'm, I'm having a conversation with this bot, and it's time to pay, and now it's gonna ask me, "Oh, you know, here's 18 different ways to pay. Which one do you want?" That, that, you know, almost, like, gets rid of the convenience factor to have so many payment options. Yeah, potentially. And I, and I, I don't think it's clear how it's all gonna play out at this point. But it's, it's an interesting evolution to start getting some real mobile wallets on, on Facebook. Um, the There was also an event within the event. Uh, Jason Del Rey of Recode hosts, hosts a, uh, a series of commerce, um, panels at some of these events.
20:14 So I think you and I both attended the, the Shop Talk one, and he had a second one at Money 2020. And, uh, one, one of the, the panelists, um, was responsible for, for, uh, the, the marketplace at Facebook. And so we asked her if we thought that, you know, they would be adding, uh, actual transactions and payments, and if potentially this, this PayPal announcement could, could come into play for the, the Facebook marketplace. And, uh, uh, she was a little evasive on that, on that question. She, she did a pretty quick pivot but did not, not seem, not… was not prepared to announce that, that Facebook was gonna get into the transaction game.
20:56 Hmm. Interesting. Did she really attribute the whole idea to the show that, as you tweeted? She did not. I made up that tweet, I'm not sh- I'm gonna say. Okay. I had a feeling that was, uh, a little hyperbole. Yeah. So, uh, just so this isn't an inside joke, uh, we, we interviewed a, a number of stakeholders from Facebook at the, the shop.org show this year, and you asked them a great question. You mentioned that, y- that, you know, there's lots of evidence that people were using the Facebook community pages to, to meet each other and sell goods informally, and you were suggesting that, gosh, there ought to be a, a formal Facebook product there. And, uh, uh, they, they, uh, gave us kind of a, a terse answer.
21:39 And then like three days later, they announced the Facebook marketplace, and it has this real prominent, you know, position on the, in the middle of the, the toolbar on the Facebook app. And, uh, it was like, you know, you, you had, uh, very correctly predicted a, a, a major new product launch from Facebook. Yeah. Yeah. Sometimes, uh, if you throw enough stuff out there, you'll, you'll, you'll hit something. That's my theory. It's just a little sad for you that you did not include that in your annual predictions, so you'll get no credit or score for that versus me. Ah, I need to go, I need to go back and look. Yeah. Maybe I can, maybe we could record another one. Sure. Redo episode 10 or whatever it was. Exactly.
22:20 Um, so in Las Vegas, I was bouncing back and forth between that show and an IBM show that was at Mandalay Bay called, uh, The World of Watson. And, uh, this apparently used to be an analytics show that, you know, a lot of the, the Coremetrics people and those folks would have, would have attended. Um, and they rebranded the show World of Watson, um, and they brought in a lot of additional stakeholders that use, uh, Watson in different vertical markets. So this became the, the mega show for all things cognitive computing, and it was… For a first-year show, it was, it was impressive. They had 17,000 people at the show. In addition to filling up Mandalay Bay, they actually, uh, took over the, the new sports arena in Las Vegas, the T-Mobile arena, and so they had the big keynotes at the arena.
23:10 And, uh, they, they were nice enough to give me, uh, VIP seating, so I got to sit in the owner's box and watch, like, the, the Ginni Rometty, I'm sure I mispronounced her name, uh, keynote, um, which was kind of fun. And, uh, I, I got to learn a lot about the evolution of Watson and how it might apply to commerce. Um, and then I did, uh, do a presentation with the, the Watson commerce product manager, where we talked about some of the, the challenges we're still facing in retail and, and, you know, getting over the hump in, in digital disruption and, and the role that cognitive computing might play in that. Cool. If you ever meet Ginni, she loves to be called Ginni from the block, so just, uh, just, you know, throw that in there when, next time you… If you ever get to meet her.
23:53 I, I have filed that away in, uh… Hopefully I'll, I'll… Right before I ask her, uh, to let me play on Augusta. The, um, one of the things I saw out of the show that was interesting is some of the things that Watson's doing in medical, um, the medical field are pretty amazing. Like, um, some of these trials they've run, it is just 30% of the time it's discovered or, or diagnosed th- 30% of things that doctors didn't diagnose because it's read so many papers that, you know, you know, it literally can hold 20,000 papers in its head, and, and a human can't come close to that. And, um, it was able to actually diagnose a little bit better than, than humans, 30% better in, in some cases because of that da- database.
24:36 Did you, did you hear about that one? I did. The, uh, healthcare is definitely one of the verticals they're really focusing on for Watson, and so there are a bunch of cool case studies. And, you know, one of the big principles in cognitive computing is sort of training the computer, right? Like, so, you know, the big thing here is you don't program Watson, you, you train Watson. You, you give it data and let it figure out how to structure the data and what kind of insights can come out of the data. And so as they've been Putting more patient data and more health data into the system, um, Watson has been able to do some pretty remarkable things. So to your point, um, in many cases it can more accurately do diagnosis of a particular, um, patient's symptoms.
25:24 But the, the potentially even cooler thing is, uh, for patients with chronic problems, like say asthma, uh, Watson can very accurately predict when you're gonna have an asthma attack way in advance. And so, you know, they're talking about these, th- you know, this near future when you, you could have six hours warning that you were gonna have an asthma attack, and that would let you take some preventative medicines. That would, there, that would enable a whole host of things that could make people's lives better. Hmm. And then what about the commerce vertical? Any, uh, interesting news there or, or use cases? Yeah. So the use cases in commerce, um, are both… There's some, some clever ones that you might not have thought of, and there are the ones you certainly would have thought of.
26:07 So the, um, they, they certainly have natural language processing and, uh, uh, text-to-speech and, and, uh, voice recognition, right? So in terms of user interfaces, your ability to talk to, to, uh, an agent that can interpret your natural language or, you know, the ability to, to give you a, a speech user interface, um, are all capabilities in the Watson stack. And for example, General Motors announced that they were, they were using Watson as the, the front end for OnStar in all of their new vehicles. Um, so that's not revolutionary stuff to me, but that's, that's interesting to be able to make some better user interfaces, and particularly natural language user interfaces.
26:52 Um, Watson certainly is sort of a next generation product recommendations engine. So it, you know, it can be trained to make product recommendations very similar to all the, the e-commerce, um, dedicated sort of recommendations engines out there, the Sertonas and Baynotes and Rich Relevances of the world. Um, and the, you know, the IBM folks would, would argue that, that with the power of cognitive computing, that Ama- that Watson can be much more personalized and accurate than, than those earlier technologies. Um, and those are kind of the traditional Uh, use cases you think of. But what's really more interesting is the ability of Watson to structure and tag data.
27:36 So you take all the product images that you have on an e-commerce site, and you let Watson look at all those and add a bunch of meta tags to those pictures, right? Like these are ones that have women in them. These are ones that have men in them. These are, um, different size models, for example. Um, there are… You know, you can hand all the unstructured product descriptions to Watson and have Watson convert all the, that unstructured data into attributes. Um, so, you know, somewhere in that Cole Haan product description, it said that this was, uh, eight, you know, that this shoe had eight eyelets for laces, right? Like no, you know, none of the websites tell you that the shoe requires 54-inch laces, but Watson can figure that out.
28:20 And so by, um, being able to use Watson to better structure all our product data, we can enable much richer, um, shopping experiences, and we can have, you know, more accurate facets and more accurate searches and better discoverability. Um, and then we can start filtering those images based on, on ones that would be most relevant to me. So, you know, they… You can make it much easier to show, um, the apparel item I'm looking at on models that are, that are more similar to me. Mm-hmm. Not that anyone would ever photograph a, a apparel in a model that looked like me, but… What, what's the delivery mechanism for this kind of thing if listeners wanna give it a spin?
29:02 Is it a, uh… Do they have to be on web… Is it like a heavy- Yeah … full stack thing? Like you're on WebSphere, and it's gonna be a feature in five years, or are these like APIs, and you could, like a dev team could play around with them today? Yeah. So that's the biggest takeaway from this whole show is when, uh, Watson was launched, which was now Seven years ago, five years ago on that Jeopardy, um, show, the, the vision at IBM was that this was a dedicated appliance that had proprietary hardware and software and, uh, you know, that, that only, you know, mega industries would buy one of these appliances and have a huge competitive advantage in their industry by getting access to this cognitive computing. Um, and they've done a complete 180.
29:44 So today, Watson is a set of microservices that are all available on the IBM Cloud, and IBM has a, a, a portal for microservices that they call Bluemix. Um, and so you can literally, uh, go to that portal, you can create an account, enter a credit card, and you can start, you know, buying access to any of these APIs. And so all of the things I just described to you are individual little services, um, that, that you can use instantly, and they all have, you know, sample code. Um, they, uh, in many cases, they have, uh, like pre-built demos so that you can try the experience. So you could go there right now and, uh, give it your Twitter feed, and it'll read all your tweets and, and give you a sentiment analysis of your, of all your tweets, or you can upload some images and it'll tell, you know, it'll show you how it would classify those images.
30:38 And, um, and so it-- They made it really accessible and, you know, frankly inexpensive. Um, and so that's to me, uh, the most interesting part of this. Like in the, in the old world, they would've said like, "Oh, uh, we, we-- IBM WebSphere has Watson built in for product recommendations, and therefore WebSphere has better product recommendations than SAP Hybris." Um, and the reality is if, if Watson comes up with the best product recommendations, it would be extremely easy for SAP to Sign up for one of those APIs and build that product recommendation right into, to Hybris.
31:18 Hmm. Um- And is it a metered ki- metered API kinda thing? So it's like, you know, X cents per call, or is it- Yeah, that- Yeah for most of the APIs, it's, it's based on number of calls. And so, you know, but it is, it is pretty inexpensive. Like, you're, you know, you're talking about, you know, um, single dollars for tens of thousands of API calls or things like that. Yeah. Um, and it varies- Awesome … varies a little bit based on the service. But I'll put a link in. You don't have to have an IBM account, um, to go explore the portal and see kinda the, the 19 core Watson services. Um, and, you know, they're basically just putting those services together, mixing them in different ways for different industries. And so at one of the keynotes, they showed a pretty cool use case.
32:02 Um, they brought out an, a, a programmer from an insurance adjuster, and, like, literally on the stage in the keynote, he brought a database of drone photographs of undamaged rooftops, and he, he uploaded, like, you know, 5,000 pictures of, of undamaged roofs to Watson, and then he brought about, uh, 5,000 pictures of roofs that had been damaged by, um, by storms. And the, the actual insurance adjuster's estimate for repairing that roof, and so he, he basically trained those two data sets, and what, you know, he, he got back in eight minutes was, like, an API he was in- able to embed in his little adjuster software, um, that essentially let him take a picture from a drone and do a real-time estimate of the, the repair cof- cost of the roof, which was pretty cool.
32:57 Wow, two of my favorite things, cognitive computing and drones. Exactly. Uh, we, we asked him to w- to, if he would fly the… 'Cause he brought one of these commercial drones, if he would fly it around, and he, he, he mentioned something about the liability insurance in the T-Mobile arena. Yeah. Yeah, I've tried that. It's, it's not a trivial thing to get approval for. No, apparently, yeah. The, even rougher than the FAA are the, like, unions at the sports arenas. Yeah, yeah. I can imagine. Cool. Any other highlights from IBM? Um, nope. I think those were the big thing. I mean, the gimmick thing, there's, there's this robot that we've seen used in a bunch of retail locations called Pepper. Mm-hmm. Uh, and I think you met Pepper at NRF, and, uh, Pepper's used for customer service applications and all these other things.
33:40 And, uh, of course, IBM has built all of those cognitive features in, into Pepper to have a, a, you know, an, an extra smart version that can learn as she talks to you. Is this the little one that's, like, about four, three feet tall, and it, it, uh, it's, it's humanoid in shape and stuff like that? Exactly. And it has a little light. Yeah. The thing that's weird about it, the last one I saw is it had these little, um, you know, human gestures, and, like, the speaker was talking, and then it would voice activate. But then when he wasn't talking to it, it would just kind of like move around and, and like put its hands on its hips and stuff, and that was actually got a little creepy. And, like, several people I talked to were, like, creeped out by the fact that it was just kind of like, you know, making human motions there while the guy was talking.
34:23 Yeah. Did, did this, did this one … Did it have that kind of a, a, you know, little inflections? Uh, so some of them did, and it also had this, you know You can see the, the potential power of these things, but you can also see, like, some of the, the challenges of the nuances, right? Like, so one of the use cases is, you know, Pepper recommending sizes in garments that fit you, right? And that's a, you know, a potentially pretty sensitive use case when you're talking to a woman about, like, what, what size dress she should wear. Mm. And Pepper doesn't really have any of those nuances. Yeah. You have very large feet. You should wear a size 13. Exactly. Do you ski with those feet?
35:04 Yep. Uh, oh, you know, you did just remind me, it… One other, like, interesting use case to me is, um, some of the IBM product managers envision in retail that one of the most common user interfaces for these Watson APIs could be a voice user interface to sales associates. So they imagine a sales associate walking around with a headset and a mic, and when they're, you know, assisting a customer and they wanna do an inventory check, they just say, you know, like, "Watson," you know, "do we have the, the Steve Madden ten and a halves in stock?" And, and, uh, you know, Watson could do all those sorts of things, and Watson could actually be listening to the interaction between the sales associate and the, the customer, and using that conversation that Watson is hearing to start refining product recommendations and, and making custom recommendations for that in-store shopper based on their shopping history and all the data that, that Watson has access to in the, in that retailer's ecosystem, and the cognitive computing recommendations.
36:05 Cool. And do you find… I've never talked to Watson oth- other than seeing it on Jeopardy. Do you find it's Alexa level quality of understanding, or is it… it's significantly better, or hard to tell? Yeah. So, uh, it, it feels like the language recognition is probably better than Alexa. Um, the one thing that, that's really cool, and the, the one in my room just turned on. Um, the one thing that's really cool about the Amazon product that I'm not sure IBM is, is completely, is even trying to demonstrate is the sort of far field, um, voice discrimination. So the, the ability to recognize a voice from the background noise in a room, um, from a microphone that's kinda far away from the speaker, I think works really well in the Amazon product.
36:51 And, you know, Amazon, uh, IBM hasn't been demoing that technology. Yeah. And, um, I guess since it's in the cloud, they could have a form factor that could be just, like, a floating speaker kinda thing too. Um, uh, I'm not as familiar with their product line. Is that something they're doing or have announced, or you don't think that's really their sweet spot? No, I, I haven't seen any- anything along those lines. I think, you know, what they would say is, "Hey, we've got these APIs. Uh, if, if we're better at speech recognition than Amazon, Amazon should feel free to use our APIs and make that available to all of their, uh, their customers on the Echo." Or, "If you wanna build an Echo competitor, and y- you know, you're, you're Sonos or whomever, and you wanna build speech recognition into, into your product, like, you, you should use our APIs."
37:34 But it, it doesn't seem like, like I- IBM is, is eyeing Making their own agent, you know, or frankly even having like a branded Watson personality that you talk to. Hmm. Very cool. The, uh… Any other highlights or? Uh, well, so one other thing that was kinda funny juxtaposition of the two shows, IBM did announce a new payment technologies at the IBM show called IBM Pay. Hmm. Um, and you know, arguably they, they probably should've done that across the street at Money 20/20. Uh, and at first I was kind of wor- worried that like, oh no, is IBM gonna try to make a branded wallet to compete with Apple Pay and, and those guys and, you know, what, what's gonna be their competitive differentiation in that space?
38:18 And that's not really what it is. Essentially what it is, is it's the, uh, a white labeled payment API. So you're building an e-commerce site or an app and you wanna take a credit card. Um, until now if you were using WebSphere Commerce, you would go buy a payment, um, solution from a third party, most often like Cyber Source or First Data, or if you were a little smaller maybe you'd, you'd use Stripe. And so IBM is now offering their own solution in that space. Got it. Interesting, cool. So kind of a gateway thing. Yeah. Yeah. So it'll be, you know, with tokenization and fraud detection and, and all the, the bells and whistles that you- you'd typically expect from a modern payment system.
38:58 Is there a Watson tie-in, like he's doing the fraud piece or anything like that? Uh, not strongly implied in this first announcement, but like yeah, you could certainly imagine that that was possible. Cool. One other news thing I wanted to just kinda throw out there. Um, Deloitte does this holiday survey of consumer preferences around, um, retail and e-commerce. Um, pretty standard stuff. There was one data point I thought was interesting. So they asked consumers, "What do you consider fast shipping?" And they give you a couple ranges, like one hour, you know, two to f- eight hours, one to two days, three to five days, that kind of thing. Um, and what was really interesting is there's a-- one of the biggest moves-- And they-- What's, what's good about the survey is they ask it every year, so it gives you some kind of trend data in directionally what's happening.
39:46 Um, and it'll probably come as no surprise to listeners, um, the, the three-to-five-day shipping bucket last year, 42% of the-- 63% of the folks said that was pretty fast. And this year, only 42% said it was remotely fast. So, so effectively what that survey is telling you is that consumer perce-perception is essentially at, it's gotta be two days or less to be fast shipping. Um, you know, I think this is one of the things that, that was really hard for the flash sale sites. You know, they're, they gave all these discounts, but then it was, you know, seven to 14-day kinda shipping. Um, uh, uh, you know, uh, you, you and I have talked about Wish before.
40:27 Some of these direct-from-China things sound really good, but, you know, you order these, uh, you know, these widgets that normally cost $50, and you can get them for eight, but then it's-- you know, you see the shipping, and it's a four-week shipping window kind of thing. So, um, it-- That, that's something for listeners to really pay attention to is this, this shipping war i-is on. Amazon has set expecta-expectations at two days. They're pushing lower on that. Um, on this conference call, they talked about, you know, getting as many things delivered in, in hours versus days as they can is a huge investment layer for them. So, uh, definitely interesting kind of data point, and we'll put a link to it in the show notes. Yeah. And I, I really like tie that data point to, uh, with, with another data point that Comscore gives us every year in their State of the Internet, right?
41:15 Like, so clearly we're seeing customer expectations largely influenced by Amazon are, are, uh, expecting faster shipping. And then this-- the annual Comscore study shows us that every year the percentage of all e-commerce sales that happen with free shipping goes up Um, and so we're, you know, a few years ago it was about half of all, all e-commerce spending happened with free shipping, and that could be either because the retailer always offers free shipping or the retailer offers a free shipping threshold and, and 50% of the consumers, you know, get over that threshold, or maybe they have free shipping promotions around holiday. Um, but the mix keeps moving up. So now we're, we're at something like over 60% of all e-commerce is with free shipping.
42:00 And so, you know, if you're going to be in the e-commerce business or if that… or if direct to consumer is gonna be one of the channels that you expect to serve, you know, you really need to be thinking about a future in which you have to, uh, give the shipping away for free and it has to be two days or better. Yeah, and that, that's a real struggle. So I think, you know, a lot of retailers, um, are pretty aware of this. When I talk to all these brands and they're just thinking about getting in the shipping business, really blows their minds that, that the level is so high. And, um, you know, it makes options like Fulfillment by Amazon almost a no-brainer for a lot of people 'cause the investment level is, is really cranked way up there. Yeah, the barrier to entry otherwise would be very high.
42:42 And I, I think- Yeah … that, that calculus is exactly why I like the, the Andy Dunn from Bonobos quote that "E-commerce is a great business as long as you don't care about EBITDA." Yeah. Yep. Um, well, Scott, it has happened again. We've used a perfectly good hour of our listeners' time. Uh, but I want to thank you for educating all of us about the Q3 Amazon announcement, and, uh, I especially wanna thank all our listeners. Uh, without you, the show would not be possible. Thanks to you, Jason, for making the trek to Las Vegas, hanging out in the Vene- Venetian, uh, tooling around in a gondola and, uh, whatever else it is you do at Mandalay Bay.
43:23 We, we appreciate your hardships for the show. Uh, it, it's entirely my pleasure. Uh, so until next time, we'll wish the listeners happy e-commercing. You've been listening to The Jason and Scot Show. For all the latest news and trends on e-commerce and shopper marketing, subscribe to us in iTunes or visit www.jasonandscot.com.
Overcast
PocketCasts
Amazon Music
Audible
Youtube Music
RSS Feed