Join us for VISIONS Summit NYC  - June 11
Episode 22
January 30, 2017

NRF Series: IBM Watson with Amber Armstrong

Brian discusses Watson, Cognitive, and the challenges of applying technology to retail with Amber Armstrong, Director of Digital, Social, Influencer at IBM

<iframe height="52px" width="100%" frameborder="no" scrolling="no" seamless src=""></iframe>

this episode sponsored by

Brian: [00:00:24] Welcome to Future Commerce, the podcast about cutting edge and next generation commerce. I'm Brian Lange. And today I have Amber Armstrong from IBM Watson's commerce and influencers team. I am so excited to have you, Amber. Why don't you introduce yourself and say hi to the audience.

Amber: [00:00:42] Absolutely. Thanks so much for having me. This is really exciting. Excited to be working with you here at NRF. We've had an exciting few days and a couple more days coming with a few more kind of exciting things we have going on.

Brian: [00:00:53] Yeah.

Amber: [00:00:53] My name is Amber Armstrong. I lead digital and social for the newly announced Watson Customer Engagement.

Brian: [00:01:00] Customer Engagement. That's right.

Amber: [00:01:03] Yeah, Watson Customer Engagement. It's kind of cool. It's the combination of Watson Marketing, Watson Commerce, and Watson Supply Chain. And we really think it's our differentiator in the market... Cognitive across all three.

Brian: [00:01:17] Right. Yeah.

Amber: [00:01:18] No one else is doing that.

Brian: [00:01:19] No.

Amber: [00:01:20] Super excited.

Brian: [00:01:20] I think Salesforce is kind of claiming to do that. I don't mind saying that on the show because honestly, I have no idea what Einstein does. So as far as I know, you're the only ones that are applying across all three, especially.

Amber: [00:01:38] We are the only one across all three.

Brian: [00:01:40] Yeah. No doubt.

Amber: [00:01:41] There are people that are applying across two.

Brian: [00:01:43] Yes.

Amber: [00:01:44] No one, we believe, is applying across all three.

Brian: [00:01:47] Yeah. No, that's great. Yeah. So how did you end up in this role? Tell me a little bit about yourself.

Amber: [00:01:53] Yeah. So I have been with IBM about ten years, and I've been through a variety of roles. I've worked with business partners. I've worked in messaging. And about three years, I kind of fell into the digital space and really started in on the social business side. I moved into commerce.

Brian: [00:02:11] Yeah.

Amber: [00:02:12] Which is now Watson Customer Engagement.

Brian: [00:02:13] Yes.

Amber: [00:02:14] And I lead all of our web presence, everything that you see externally for Customer Engagement is mine, if you will. All of our social channels and really figuring out how we take this very complex story and dial it into something that can be consumed really easily on social and digital.

Brian: [00:02:35] Yeah. It's a good question. We'll taking this little bit later. But I really want to focus on what Watson is and what Watson isn't and what you can do now. And we'll give him some examples of that. And also what you see as going to be possible going into the future. And so, yeah. So to start, I think, going into NRF, here at the Big Show in New York.

Amber: [00:03:05] You can hear the ambulances outside. {laughter}

Brian: [00:03:07] Yeah. Yeah. New York ambiance. It's really loud sirens right now. So, yeah going into into the show, you guys released a really groundbreaking study on Generation Z. And one interesting thing that I noticed is that Gen Z is actually still about in-store, which is really cool. They pretty much said that they're not all about pure play eCommerce. They want to be able to go experience brands in a physical environment. However, I think they have said they want a completely different experience than the last 20, arguably 80 years, of what we've seen in in-store shopping experiences. So, yeah, maybe you could unpack that study a little bit more and talk about Gen Z a little bit.

Amber: [00:03:58] Sure, yeah. So I think the thing that's really interesting with Gen Z is personalization to them is completely, totally normal. So in the study, it says 44% of them don't watch television. They don't have a concept of this one for many marketing approach.

Brian: [00:04:16] Yeah.

Amber: [00:04:17] They're used to Facebook and Snapchat, and one of the things I like to talk about is your Facebook feed is as unique as your fingerprint, right?

Brian: [00:04:26] Yeah.

Amber: [00:04:26] There are very few things out there that are truly unique to individuals. But Facebook and any social channel is a perfect example of that. Right? Because you and I could take the same actions. We could friend the same people. Those people may not from its back. We can attempt to. But then how what you click on in your news feed, and what I click on is going to be totally different.

Brian: [00:04:48] Right.

Amber: [00:04:48] So they're used to having things that are built immediately for them.

Brian: [00:04:52] Yeah.

Amber: [00:04:53] There's no generalization for Gen Z.

Brian: [00:04:59] It's an insane concept because having grown up in the 90s, like, we didn't have that at all. Our experience was completely pushed on us a hundred percent. And I've talked about this a few times. But like my experience, for instance, like playing video games, growing up, it was all like one very specific experience that my friends and I all shared. But now when my little nephew goes to play video games, he's playing Minecraft, which he can mod to no end and do pretty much anything he wants. And I think that doesn't get to personalization. But the point is that experiences are very unlimited now.

Amber: [00:05:46] Yes.

Brian: [00:05:46] And so for brands, in order for them to actually provide value to Gen Z, if they are personalizing, they're going to lose.

Amber: [00:05:56] Well, when they go into this store, they expect to see exactly the thing that they looked at online. Because the other thing is... So they say, yes, we like to go in store and purchase things. That's great. But you know what? They also have done all the social before they get there.

Brian: [00:06:14] Right. Right.

Amber: [00:06:15] They've read the reviews. They've talked to their friends. They've done all of that research to get them there. So you've got to make sure that their experience is personalized, but also that it's linked to them across all of your experiences.

Brian: [00:06:33] Right. Like they need to be able to... It's almost like attribution. Kind of.

Amber: [00:06:38] Yeah.

Brian: [00:06:39] Yeah. That's huge. So what are some practical, high level ways that you feel like brands could do that? Or maybe you have some examples of ways that Watson has been applied to help brands to be able to do this for Gen Z.

Amber: [00:06:57] There are two, actually, that we just announced the last couple of days.

Brian: [00:06:59] Yeah. Yeah.

Amber: [00:06:59] Watson Customer Engagement. Hermès. And also HSN.

Brian: [00:07:04] Yeah.

Amber: [00:07:05] And so anything with Hermès, what they're doing is pricing optimization.

Brian: [00:07:08] Right.

Amber: [00:07:08] So they're looking at. OK. We have all of these stores... And they're the largest retailer in Cyprus. They have all of these stores. How do they optimize the prices in their stores? So it's the what and the when and the why.

Brian: [00:07:23] Yeah.

Amber: [00:07:24] They're powering all of this cognitive analytics and data behind it to make those decisions and then tying it in with the supply chain to know how much inventory do we have this thing? Is that inventory in the right location? Can we get it there? And they're able to make all those decisions happen automatically.

Brian: [00:07:43] Yeah. Yeah, yeah, definitely. So for a customer who is interacting with Hermès, and they're walking into the store, and they're seeing this pricing, what does that experience feel like? Is it something like you see prices dropping and following on the fly. Or is it something where it's like it's not actually apparent to them that anything different is actually happening?

Amber: [00:08:13] It shouldn't be apparent to them.

Brian: [00:08:14] In one of the keynotes, I think Levi's was saying one of the most common causes for abandonment of a purchase is being out of stock. And it kills merchants all the time. So being able to have that level of data and be able to make sure that things are priced appropriately, so that you're guaranteed to to be able to sell at any given time. That's huge.

Amber: [00:08:38] Well, and also tying it across the three. Think about this, and this happens all the time, as much as we would love to say it doesn't. But marketing is like, oh, this ad is performing really well. I'm going to buy more space on Facebook for this ad. We can sell this lot. That's great. And then suddenly look at all this traffic from this ad and suddenly everything is personalizing on the website. And the guy over in China who is shipping the goods is like, what is going on?

Brian: [00:09:12] How? And then there's nothing to even sell anymore.

Amber: [00:09:14] Right.

Brian: [00:09:14] So if in reality, if you had left the price at a proper spot and advertised at a proper rate, then you would actually be able to meet your demand where it's at..

Amber: [00:09:27] And even being able to take the feedback from that. To have the marketer say, hey, we're seeing a lot of really great traction on this watch. This particular color. Feed that back in through the supply chain so that the guy who was actually manufacturing the watches starts making more blue watches.

Brian: [00:09:44] Right. Yeah.

Amber: [00:09:47] So that's going to increase your profits. It's going to increase sales.

Brian: [00:09:49] Absolutely.

Amber: [00:09:50] The whole thing will flow together. The challenge is these companies don't have the ability for this data to flow between these groups. These groups tend to be very siloed.

Brian: [00:10:01] Siloed. Definitely. Oh, yeah, no doubt. I have seen that before. It does cause problems. I think I can safely say that almost every merchant out there has felt this pain.

Amber: [00:10:13] Yes. Absolutely. Well, HSN is the other customer I mentioned. They're using Watson Marketing. So they're kind of on that front end I was talking about. So how do you make sure there's the same experience, right? Whether someone's watching Home Shopping Network on television, or they're online, or mobile, or whatever it is, making sure that they're seeing the same experience that they're able to participate appropriately.

Brian: [00:10:37] Yeah. Yeah. Definitely for all of you out there that have really invested your lives into responsive design, this is kind of like making your brand responsive across all channels. So it's really amazing.

Amber: [00:10:49] That's a great way to put it. It really is making sure that that exact same experience shows up no matter where someone is.

Brian: [00:10:57] Yeah, that's something that I think little merchants, again, wish that they could do and would have the confidence that that's actually happening. I think right now they don't even necessarily know half the time. So, yeah, it's great. Cool. I mean, I think one of the things that I talked about this earlier, the power of Watson, it almost seems limitless. I mean, it's AI. It's such a broad idea in itself. And so one of the things that we've talked about on the show before, and we had a Watson SME from your partner network come on the show before. And he talked about some of things that Watson is able to do. But it'd be really great to kind of hear that reiterated again. How would how would a retailer or merchant best go about actually applying Watson to their business? How can they actually engage with IBM to say, OK, yeah, this is what I need from Watson and this is what Watson can do for me. What does that experience look like?

Amber: [00:12:04] Well, I think it really is limitless at this point. If you think about where we are, and this is one of the things we were talking about yesterday. There're so many things. We're at such a turning point in the world of commerce and marketing and supply chain.

Brian: [00:12:19] Yeah. It is.

Amber: [00:12:20] We don't even know what's possible, yet. We haven't even dreamed it.

Brian: [00:12:23] Yeah.

Amber: [00:12:23] There are things that you can do today, and there are some great customers who are. Starting out, I think one of the things that we talk a lot about at IBM is this Agile Methodology.

Brian: [00:12:37] Yes, I love Agile. But I just wrote an article about Agile recently.

Amber: [00:12:41] Oh awesome. Great. It's about taking little steps and tweaking as you go.

Brian: [00:12:48] Iterating.

Amber: [00:12:49] If you think of using Watson as this, I'm gonna go do this giant thing that's going to be four years out, your competitors are going to beat you to it because they're going to start doing little things, and they're going to learn as they go, and they're going to get there so much more quickly. And there're a lot of different ways to do it from the marketing side and the commerce side. And the people are really surprised to hear that we have at Watson Content Hub a 30 day free trial, $250 a month.

Brian: [00:13:24] Which is mind blowing. Actually, I didn't even really know about the Content Hub. When did you release this?

Amber: [00:13:29] Maybe a month ago.

Brian: [00:13:30] Maybe. Yeah, that's what I thought.

Amber: [00:13:31] Maybe two months. I think someone was saying November. Yeah. And you don't have to say this now if you don't want to. But this is direct competition for like Adobe Experience Manager.

Amber: [00:13:46] Definitely.

Brian: [00:13:46] Right. So $250 a month is crazy for something like this. And I know this because I've lived in this world. It's an unbelievable price to be able to engage with AI applied to marketing. That's mind blowing. So what you're saying is to kind of sum up what you're saying is find something that Watson does that does a lot of things that you know that you can leverage out of the gate and start there. And then as you start to experience the benefit of that, then maybe start to build up, or sort of build up some sort of MVP. So it's like, okay, I can see value in these things immediately. I know it. It's guaranteed. And then build on top of that. That's great. Instead of saying oh, I want to go do something crazy with AI and hey IBM, what can you do for me? Engage at a smaller little first and get live.

Amber: [00:15:02] One of the things I love about Facebook and what Mark Zuckerberg does there is he does, and lots of companies do, this kind of...where he challenges people to get locked in a hotel for the weekend. That sort of thing. I mean, why not do that? Take and lock a bunch of your developers in a hotel for a weekend, and turn loose on the Watson APIs. They are free. There's conversations. There's the retrieval reword and personality insights. We've done some cool things with it. We did at something called the Cogni-T-shirt Company that we took to a bunch of different events.

Brian: [00:15:43] Oh I love that.

Amber: [00:15:43] And we use people's social profiles. This is just us kind of playing around with it. This wasn't like some massive development. But we played around with it. We used tone analyzer and all these different Watson APIs to take people's social profiles, plug them in, ask them some other additional questions on top of that, and then customize a t shirt for them. It's a very simple use of the technology. It goes to show you if you're a retailer, if you use some of these APIs combining your customers' social profiles, maybe ask him some questions that also gets their buy in, their participation, and then customize something for them in-store. That's a very simple, simple, simple example that makes it real for your customers and also can help your team learn about cognitive and the process.

Brian: [00:16:35] I love it. Yeah. One of the things I wanted to ask you, and I meant to ask you kind of was a follow up to my last question, is Watson can be applied sort of within the context of IBM only solutions, obviously, right? But can you apply Watson to other solutions as well?

Amber: [00:16:56] Yes.

Brian: [00:16:56] You said you mentioned the API, but so let's say I had a Magento store, and I wanted to apply Watson to pricing strategies. That would be a perfectly legitimate use case.

Amber: [00:17:10] So there's a couple of different ways we think of Watson. So when you think of something like Watson Content Hub. Watson and cognitive is infused in that solution. So when you buy that solution you are getting Watson. Then there's also this platform as a service idea. So this Watson API that you can access. Conversations. The personality insghts. All the things. Tone analyzer. Those things we were just talking about. Those are as a platform. You can do whatever you want to with those. Our business partners do that as well. But we also are making sure in order to get the Watson Customer Engagement, be able to use the Watson brand with that, we had and how Watson infused in our solutions.

Brian: [00:17:50] Right.

Amber: [00:17:51] So it's both.

Brian: [00:17:51] It's both. Makes us. That's kind of what I was driving at. That's amazing. So, in short, I think this is something that anyone can access and take advantage of. That's really what you're driving at. At least in some way, even if it's a small step.

Amber: [00:18:12] Let's play. Play a little bit.

Brian: [00:18:15] Play a little bit.

Amber: [00:18:15] Encourage innovation and have people dream big and start small.

Brian: [00:18:20] So Watson is for the entrepreneurs. For the innovators.

Amber: [00:18:24] Absolutely. Absolutely. This is I mean, it is certainly applicable at a big company. It's also very applicable for any company. And that's where, especially sitting here at NRF, these retailers, especially some of the really big retailers, their concern shouldn't be their peers. Their concern should be the companies that are significantly smaller than they are, that are so much more nimble, that are trying things out and using new technologies and maybe not beholden to all of this kind of on premise existing legacy type platforms.

Brian: [00:19:03] Right.

Amber: [00:19:03] They've got to think like a challenger brand.

Brian: [00:19:05] Yeah. We actually heard Richard Branson talk about that this morning in the keynote. He was talking about how to be competitive. And I think he was pretty much was giving at this very idea, which is the small guy is the guy you got to watch out for. If they're innovative and they're cheeky, as he put it, and they've got spunk, they are going to come up and get you when you aren't looking.

Amber: [00:19:35] Absolutely.

Brian: [00:19:35] So you have to be the same way back. And it's one thing I think enterprise little businesses sometimes miss is in order to compete, you also have to be innovative. You have to be out there, and you have to be providing solutions that your customers are expecting.

Amber: [00:19:56] Absolutely.

Brian: [00:19:57] Yeah.

Amber: [00:19:58] And it's it's about predicting their needs. How do you get ahead of where they're headed?

Brian: [00:20:03] Yeah, totally. Speaking of getting out of ahead of things, I think based on the volume of voice driven tech that we saw released at CES and there was some of that here at NRF, it really looks like 2017 is going to be kind of the first really dominant year of voice interaction. And I say that, I mean, last year was it was strong, no doubt. But Amazon sold twenty six million Alexa enabled device devices in November. Which is crazy. Twenty six million homes are now, well I guess it wouldn't necessarily be homes, but twenty six million devices are out there that can be talked to and just for the month of November.

Amber: [00:20:51] Yeah. {laughter}

Brian: [00:20:53] And at CES we saw Alexa in refrigerators and robots... Watson's in a robot. We saw Watson and Pepper here at NRF. And so looking forward, looking out ahead of the curve, what do you have to say about voice? I mean, how do you feel like consumers are going to leverage voice for purchases specifically?

Amber: [00:21:23] Yeah, yeah. Well, I think it's really interesting because you think about the Amazon work, and you take the work that Mark Zuckerberg did over the last year for his work. He was automating his home, right?

Brian: [00:21:35] Right. His personal assistant. Morgan Freeman was narrating.

Amber: [00:21:40] Yeah. Really great. He found he actually used text more than voice in that scenario. So I think you have to have both.

Brian: [00:21:49] Both. Absolutely.

Amber: [00:21:50] And the difference is when someone writes something, they write in one way. When someone speaks something, they use natural language, which causes this totally different interaction.

Brian: [00:22:01] Absolutely.

Amber: [00:22:02] The system has to be able to understand both. So we had a couple of really good examples of customers that are doing us today.

Brian: [00:22:10] Oh, great. Yeah.

Amber: [00:22:11] Yeah. So 1-800-Flowers.

Brian: [00:22:13] Actually we've talked about this on the show a little bit before. Gwyn is so cool. But not in detail. So yeah, dig into it.

Amber: [00:22:21] Gwyn is great. Gwyn is "Gifts When You Need." That's what it stands for, so I think that's kind of cool. And you go in and you interact and Gwyn asks you questions. So it's using the Watson Question and Answer API. And so if you ask Gwyn questions about whatever you're trying to do, Gwyn will ask you questions. You kind of have this back and forth.

Brian: [00:22:41] Right. Conversational commerce. Right?

Amber: [00:22:46] And so then it recommends solutions to you for whatever gift you're trying to give. So then you think about it's also learning as it goes. So the next time you come in, it's going to know more about you. It's also going to know. Well, hey, I seen this interaction happen across 10 customers. So it's learning from that. It's getting smarter and smarter as it goes.

Brian: [00:23:11] Amazing. I think this study that came out on this, did it say something like 80% of customers that that they loved the interaction and that they would definitely use it again.

Amber: [00:23:21] Yeah. And I think the trick to that also is knowing when to switch. So we talk about when Watson is the combination of man and machine. It's not just machine. It's the combination.

Brian: [00:23:37] Yeah. So does Gwyn switch at some point?

Amber: [00:23:40] Yes. If it needs to.

Brian: [00:23:41] Fantastic.

Amber: [00:23:41] At some point it recognizes that oh hey, I am a little over my head here. I need to get some help. I need a human to help.

Brian: [00:23:51] Yeah. We talked to some other vendors that are using the exact same approach to conversational commerce. Because the truth is, you know, at some point you really do need that level of human touch to finalize things. Although I think that even the Watson team is probably working to get even better to the point where maybe humans don't have to touch things in the future.

Amber: [00:24:16] Certainly that's the plan. It's this progressive the human has to get involved later and later. So that you're investing your human resources and the things are going to make those experiences even better.

Brian: [00:24:31] Yeah. Totally.

Amber: [00:24:32] So you're not spending the time answering the basic questions with a human.

Brian: [00:24:37] Yeah, that's great. What about in-store? I think as we get further down the road with voice, I can't help but hypothesize that we're going to start interacting via voice, even in-store with AI. Do you have any good examples of this happening?

Amber: [00:25:00] Yes. Yes. It's starting. It's really starting to get out there. South by Southwest is coming up in a few months. We did a big experiment at South by Southwest last year where we had lots of robots in our environment, and they were all interacting. People were interacting with them all using voice.

Brian: [00:25:21] So cool.

Amber: [00:25:21] We even had one which I don't even understand how this happened. You could put this thing on your head and tell the little Star Trek robot... Or was it Star Wars? I get them so confused. I am so sorry.

Brian: [00:25:36] {laughter}

Amber: [00:25:36] Move it around. So there's all kinds of things that that that you can do when you're in a physical space.

Brian: [00:25:45] Yeah.

Amber: [00:25:47] One of the things we've seen customers starting to play around with is bank tellers. So you walk up, you have a conversation. Here's why I need to do. And the robots can talk to you. SoftBank is the company that's working with this. There's also a company called Rabobank. And their whole idea...

Brian: [00:26:07] Was that rob a bank? {laughter}

Amber: [00:26:12] Similar. Similar. It's Rabob. {laughter}

Brian: [00:26:14] Ok, good to know.

Amber: [00:26:16] It's great because their whole idea is around how do you create automated banks that really have these robots?

Brian: [00:26:27] Is that released?

Amber: [00:26:27] It's out there. You can go look on their web site and read all about what they're doing. They're still learning.

Brian: [00:26:32] So cool. Yeah, definitely going to go check that out.

Amber: [00:26:36] So there are a lot of people who are experimenting. That's what we have to do. I also think that that really helps retailers give their customers to embrace some of the novelty of it. So it goes a little bit from there's this kind of creepy thing to it. I mean, Pepper is really cute...

Brian: [00:27:02] I took a picture with Pepper today, and I sent it to my family and my boys thought it was the best thing ever.

Amber: [00:27:09] Yes. Back to Gen Z, right?

Brian: [00:27:12] Right? Yeah.

Amber: [00:27:13] How do you engage with Gen Z? They're going to love to talk to robots, I bet.

Brian: [00:27:17] Oh, for sure. Oh yeah. My kids would love it if we had a robot they could talk to. They would boss around. Yeah. That's awesome. Yeah it's really exciting. So what do you see if you could look out into the future a little bit... What do you kind of see some politicians looking like? Like if you start getting creative a little bit and apply AI to the in-store aisle, what would that look like to you?

Amber: [00:27:46] I'm not sure about the in-store aisle, but I know on digital I can tell you how I'm planning to use it.

Brian: [00:27:54] That's great.

Amber: [00:27:55] So as I was saying earlier, all of our digital experiences, one of the sites is Think Marketing. And what I'm gonna be able to do with Think Marketing... Think Marketing is a new site specifically for marketers. It's also interesting to commerce leaders that are trying to talk about customer experience. And when you go in to Think Marketing right now, you're reading an article on the right hand side it's Watson Recommends. And it pulls up articles. It pulls up products and pulls up customer studies. All these things based on, we're not programing it, it's based on the content and the article and understanding that natural language. And then where we're taking that is being able to take third party data. So think of LinkedIn and Facebook, what we call universal behavior exchange. Any demand based, any kind of third party you can think of being able to consume that. So that when you come in to Think Marketing, Think Marketing is customized based on your past search history, meaning search history on our site. Plus this third party data. So that's more and more relevant for you. But then I think also giving you opportunities to personalize it even more through choice. And so to make that experience something that's really unique for every single person that visits our site, we're working on Think Marketing right now with some great things coming up before Amplify here at the end of March. But then doing that across all of our sites. And thinking about how retailers can do that as well. Third party data is only getting better and better and better.

Brian: [00:29:32] Right. It really is.

Amber: [00:29:35] It is a little scary.

Brian: [00:29:35] It is a little scary. Yeah, definitely agree.

Amber: [00:29:37] But if you can use that to make the customer experience better and do it in a way that's not intrusive, and give customers an opportunity, as I said earlier, to participate in that personalization, that makes that... If I give you permission I'm much less offended when you keep following me around the booths.

Brian: [00:29:57] Absolutely. And the truth is, I talked about this with Ryan Eisenberg as well, basically once utility overwhelms concern, we as humans just jump on it because we're all about things that are actually useful, things that that are going to benefit us. And so if we have a brand that we trust that we're giving permission to then we will be willing to do that. And then Amazon's good example of that.

Amber: [00:30:31] Absolutely.

Brian: [00:30:31] So, yeah, anyway, I think we're well over time at this point. Thank you so much for being on the show, Amber. It was really great talking with you and I'm excited about what's ahead with Watson for sure.

Amber: [00:30:43] Great. Thank you so much for having me here and thanks for talking with us at NRF.

Brian: [00:30:46] Absolutely.

Recent episodes

By clicking “Accept All Cookies”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. View our Privacy Policy for more information.