Latest / Talk Commerce / The Importance of Analytics in AI and Machine Learning with Lori Schafer

Transcript

The Importance of Analytics in AI and Machine Learning with Lori Schafer · Talk Commerce

• Brent Peterson (00:02.229) Welcome to this episode of Talk Commerce. Today I have Lori Schaeffer. She is the CEO of Digital Wave Technology. Lori, go ahead, do an introduction for yourself. Tell us your day -to -day role and one of your passions in life. Lori Schafer (00:17.71) Okay, well, Brent, thank you very much and thank you Talk Commerce. My name is Lori Schaeffer and I'm the CEO of Digital Wave Technology. And Digital Wave Technology, the easiest way to think about it is all things product. And we will talk a little bit about what all that means, but basically we combine a master data about everything about a product and we're able to very easily and it's connect into the digital commerce as the, I'll say the platform that drives everything on that site and also a lot of AI and generative AI content. So that's what we'll be talking about today. Passions, I have a lot of passions, Brent. So it's hard to pick one, but since we were just, I think as I got on, we were talking a little bit about weather and, The fact that I live in Florida, I've always had a passion for the ocean and understanding sea life and the beach and sports within the ocean, paddle boarding in the ocean, surfing. I'll pick that one for today. Brent Peterson (01:35.861) That's a that is a fantastic passion. We're we're both connected with water. Minnesota has lots of water. I think there's an there's as many boats in Minnesota as there are in Florida, and we probably have an equal amount of shoreline if you can believe it. But anyways, let's we'll move on. Before we get started, though, you have graciously volunteered to be part of the free joke project. I'm going to tell you a joke. And all you have to do is give me a rating on that joke one through five. Lori Schafer (01:42.624) A lot of likes. Lori Schafer (01:50.126) I can believe it. Lori Schafer (02:05.23) I think I can handle that one. Brent Peterson (02:06.293) So yeah, and I have a very good one today. You're just gonna love it. All right, here here go. Why are chickens always so mean to each other? Because they use fowl fowl Lori Schafer (02:22.542) Very good. I would have to give that one a four and a half. Can I do a half? Yes. I was waiting for it to cross the street, but that didn't happen. Brent Peterson (02:24.213) Yes. Brent Peterson (02:30.037) wow. All right. Yeah, absolutely. Yeah. Do you like how I kind of ramp that up? Like, this is going to be a great joke. And then I just landed it with, my gosh, you're, you were just like, yeah, I saw the eyes roll. all right. So products and marketing, and I love this subject. I've been, I was in the key e -commerce space for a long time as an agency owner in the E in the digital solution integrator space for Magento specifically, but we didn't really ever. Lori Schafer (02:51.086) I saw that. Brent Peterson (02:59.029) our role was never to get their product sold. Our role was to get their product to be presented to sell. So I'm always fascinated after that product is there or as it's getting there, tell us a little bit about how you help those clients in that part of it after they've got their product. Maybe you even help with the development of the product branding and things like that, but tell us, give us some background on how that plays in. Lori Schafer (03:27.47) Sure, so a little bit of background on what digital wave does is again, as I said, all things product. It's really, when I think about it, it's all things about getting that product to market from a technology standpoint. So one of the big areas that we cover is product information management, master data management, and really that foundational data lake, if you will. of everything about a product, not just what you need from a standpoint of product content, but also all the historical sales and the projections and anything to do with the financials. So really think about it as literally everything about a product. We can even get in the attribution for product lifecycle management and a number of other things. So we then have a number of solutions that are used on top of that. Think about that product information, product experience infrastructure, everything from a lot of AI solutions that do everything from automating product content, automating attribution, making sure it's the best attributes for SEO, things like planning, assortment planning, financial planning, pricing, dynamic pricing. Lifecycle pricing so a number of different things that all use that same product content Brent Peterson (04:59.477) Yeah, that's interesting. So I've had quite a few PIMs on and that's product information management systems. How do you recommend a client get into the different pieces of that puzzle as you get the product to market? So typically you'd go, you know, you develop the product, you put the product in your PIM and then your PIM would pick the channel to go to. It could be your e -commerce, your POS or... Lori Schafer (05:04.558) Yeah, very different. Brent Peterson (05:27.189) You might have multiple stores or multiple languages. So tell us a little bit about your philosophy behind that. Lori Schafer (05:34.35) Sure, it's a great question. So one of the things we did, Brent, and I've been in the retail or retail tech industry for 30 some years. So know a lot about merchandising in general and marketing in general. Ecom is a piece of that. Certainly now it's really, when you really think it's ubiquitous shopping. And so when I, in the team, we started the company, we look back at the other. providers out there, master data providers, PIM providers, and we looked very quickly at, okay, what is it that we need to do radically differently to really stand out? And one of the things is full automation, because so much of e -commerce is not automated. I think back to the days I used Magento, now these were a while back, but you're hand keying in the product. description, you were hand keying in the attributes, you were hand keying in, you know, we were trying to take photos and bring the images in. And what we've done is really looked at it more as a foundational recognition that the world is going towards AI, it's going towards analytics and traditional other traditional companies of MDM and. PIM, for example, were built at least, say, five years back. And they weren't built with the thought of analytical analysis of everything that you can do to automate and analyze the products right within that platform. So we are, the difference with Digital Wave is everything is AI native and it's on what's called an analytical database. that can number crunch high amounts of volume. Because what we recognize is the ultimate end users, they want to get answers fast, they want to get, they want to bring product to market very quickly. And then once the product is to market, and you've got that speed to market element done, you want to start to incorporate a lot of the consumer data and have that all automated into the e -commerce process. Brent Peterson (07:56.789) Yes, so tell us, do you sit between the PIM or the e -commerce platform or you sit aside it and then give them advice or analytical advice, I should say, to help manage that process? Lori Schafer (08:14.158) It's a great question. So the way the architecture that we have is designed is that we can, it's mock architecture machine, with all built -in machine learning, and we can pretty much drive the hierarchies and all of the automation that then drives right into that e -commerce as the front end of it. So that's how that works. But it's really about back to, you know, I think what you want to talk about on the call less about the technology and more about, okay, so we're kind of, you know, why do we care about all this? And when you look at the way the consumer's changing and the way technology is changing, and the fact is that, you know, companies have to become far more automated and more efficient with the mundane tasks so that their teams can focus more on, you know, how to, you know, better bring their products to life and sell more for, you know, to the consumers. Brent Peterson (09:17.621) Yeah, you mentioned earlier about having to hand key things into Magento and make sure that the products there. What are some of those mundane tasks that you see trending in this next year now that are going to be solved by solutions like you have? Lori Schafer (09:32.11) Yeah, so great question. So I mean, everybody is hearing generative AI until you're blue in the face and sick of hearing it from a marketing standpoint. But what I want to talk about more is, okay, things that are actually working for retailers and brands right now. And some of that is, again, I'll call it the very early steps of generative AI, meaning that, you know, we all know this is going to... change the world in a lot of ways. But in terms of the e -commerce space right now, certainly a couple, if I think of different, I'll call them use cases or solutions that we're quite familiar with and still brings tremendous value, both from an efficiency standpoint as well as profitability is, let's take for example, probably the most mundane but done right. being able to produce all that product content in BrandVoice. And using AI techniques that are able to batch process hundreds and hundreds of items and be able to write that content so that it sounds just as though a human wrote that content in the BrandVoice that that content should be written in. So ensuring that certain words are not used. ensuring that it's, you know, if you speak in first person or third person that that's used, ensuring that certain key phrases that tie back to the brand are used. And then what it allows is copywriters or people that, you know, they may not be called copywriters, but those in the e -commerce world that are having to in the past hand key all that in. Now what we're finding is that those individuals become the editors. They're the editors in chief and let the computer do the more mundane lifting and then they can just look at each piece of content and approve it and then be able to focus on other things. Another area is automatic attribution. And even over the past, gosh, 15 months, we've gotten extremely, Lori Schafer (11:51.15) good by working with a number of retailers and brands on being able to example, for example, you know, just by looking at the image of a product and being able to scrape off all the ingredients, for example, being able to look at a blazer and instantly be able to tell what type of lapel it is and, and you know, what color it is and you know, what type of sleeve it is. Those sorts of things have all been perfected now so that when trying to figure out those attributes first pass, a lot of that work, say 70 % of it can be done automatically. Also, when you go back in and you look at your current data, you can fill in the holes. And we've worked with companies from everything from large DIY to fashion, high fashion apparel, and really being able to streamline that whole product attribution. And that out to SEO, when you think about SEO, oftentimes the attributes you have may not, on e -commerce, may not be the ones that consumers are searching for. So being able to come back and say, have you thought about these attributes? Because that's what the searches out there are saying, and that's not perhaps what you have on your site. All of that, you know, turns into, obviously, in the long run, a happier consumer. they're going to find your products and hopefully raise conversions. Other areas like being able to instantly summarize reviews within this technology, being able to look at any product, see all the reviews across different, if you're a brand across different retailers, if you're a retailer, all of your reviews or even outside elsewhere by pulling those together. And then saying, you know, these are great comments that customers have made in the review and let's incorporate those with the press of the button right into product content. So that you're bringing that to life by here's what our consumers are telling us. These are all things that just really now with generative AI that we're able to do very quickly. I mentioned a fourth one, which is advertising, you know, above the fold, below the fold. Lori Schafer (14:15.79) being able now to combining a number of different AI models together so that we can create the lifestyle images image right, you know from from scratch that is tailored to the the type of ad you want automatically create the tagline the brand logo and Be able to create an ad which may have taken You know weeks and weeks in a matter of a couple minutes. And then you can still edit that ad. You don't have to take it as is, but it really gets the juices flowing. It uses the e -commerce team to use what only humans can do and let the computer automate the mundane tasks. Brent Peterson (15:02.037) Yeah, that's such a good description of AI. I heard an interview this last week about the found from the founders of Etsy and they were saying how maybe Etsy couldn't be reproduced now because people's shopping behaviors have changed so much in the last 10 to 15 years or 20 years almost. What other trends now do you see happening in shopping that people or that merchants are going to have to adopt? Lori Schafer (15:12.398) Yeah. Lori Schafer (15:23.082) Mm -hmm. Brent Peterson (15:30.773) to match what shoppers are doing. Lori Schafer (15:34.862) Well, obviously personalization everybody hears about, you know, that will happen. I don't think we're all quite there yet, but people do expect for the shopping experience to be personalized. And I think the technology, in some ways the technology drives the behavior and in other ways technology needs to catch up with consumer behavior. And, you know, on the personalization front, I think the technology is catching up. in the sense that now you're going to be able to, and it's doable right now, certainly, but the adoption hasn't quite caught up yet. But the idea of being able now to not just personalize products, complementary products, but also be able to personalize the text that's on the screen. So that's one area that can be really... If I look at a trend from a shopping perspective, the rise of social commerce is another one. You know, things have really dramatically changed. Where in the main social commerce sites or social media sites, that's if you look at a TikTok or you look at a, you know, a Instagram or others, you know, that's where a lot of the shopping is now happening. So it's critical to be plugging into that. Also the video, shopping on your own sites and companies that do that, that's certainly a trend that's happening. Sustainability is yet another trend that retailers and brands are going to have to ensure that on their site that they have those key attributes that tie to the sustainability of the products and also of the brand and what they stand for. So those are a few. Price sensitivity, right now with the way the economy is. Certainly that's a hot topic, a dynamic pricing. So again, I could go on, but those are some of the key things that companies are gonna have to adapt to. It's not gonna go away. Brent Peterson (17:42.901) Yeah, from a technology standpoint, then, you've mentioned AI, you've mentioned generative AI. How important is the analytical part? I think the analytical part of AI is something that a lot of people, especially end users, overlook, but maybe merchants are more concerned about, because those really drive how the system returns back to the customer. Tell us a little bit about how important the analytical part of machine learning is compared to just generative. Lori Schafer (18:13.582) Yeah, so generative AI is really, first of all, is analyzing text, voice, things that are not, as I call it, numerical AI. Meaning, and traditional machine learning AI is more around things like, you know, you think of price optimization, optimizing the size points of products. So optimizing, you know, the workflow of clicks and click -throughs and... where are consumers going on the site? I mean, the analytical side is becoming more and more important. You know, I think e -commerce, it grew up in a silo separate from the rest of the brick and mortar. And now those systems really all have to come together. And the analytical AI had been perhaps in e -commerce was more on site click through and Outside of that in brick and mortar, things like price optimization is an easy one to think of. Inventory optimization, demand forecasting, that's all AI. And now when you think about it, you really need to combine generative and I'll call it numerical AI to come up with the best answers for the individual consumers around what products do they want, what price. you know, value and also to be able, you know, to, to be able to track that inventory cross cross channel. Brent Peterson (19:50.997) Yeah, you bring up an interesting point that machine learning has actually been around for years and years and years. And now, AI is such a buzzword and, but it's been around in a different form, just like an algorithm theoretically is an AI because it does learn. From a technology standpoint, how do you see or what do you see the biggest impact if you don't say AI? Is there new technology that's coming that's going to impact how shoppers behave? Lori Schafer (19:57.39) Long time. Lori Schafer (20:09.582) That's right. Lori Schafer (20:23.342) Apps, if we don't say AI, meaning don't talk about AI and talk about everything else. Yeah, okay. Blockchain, certainly, is something that is something that when you look at tracing the life cycle of a product, I think about jewelry, for example, expensive product you want to track and understand where it came from. People are very focused on sustainability. Brent Peterson (20:27.957) Yeah. Lori Schafer (20:52.174) and that blockchain is going to give that capability to really track the ethical process of that product as well. That's one. Certainly, I guess you could technically say it's part of AI, but I'm going to say the other thing that is happening fairly quickly is voice technology. It's been around for a while. You think about Amazon and Echo and Google. And there's been voice technology out there, but it's now coming to a point where consumers are going to be expecting in the next decade to interact with the website through voice. And that's very different when it's your website. And they want to be able to ask questions to that website through voice, and they want to be able to get answers through voice. And so that's another whole area that I think is developing very quickly. Even in backend technology, being able to ask the questions and have the technology be able to sort of let the answers through voice. So that's another example of technology that is bringing brought to life. Augmented reality, virtual reality, certainly that ability to try on in the dressing room or check out your furniture and your... you know, in your house, be able to try on makeup, you know, those are other types of technologies that are now evolving very quickly. And it's the retailer or brand's job to figure out, okay, which of those are practical for the brand message that I want to communicate, you know, to my customers. Brent Peterson (22:40.181) You talked a little, you talked earlier about the ability to generate content in a catalog using the voice of the merchant. There is a voice that is recognizable now for AI and there's phrases that are repeatedly said. How do you think merchants are gonna handle, it's kind of like an accent. Lori Schafer (22:48.814) Yes. Brent Peterson (23:06.709) A lot of newsrooms that have a Midwest accent, but people sometimes want to hear a Southern accent. I think that the the Grammarlies of the world and AIs of the world are really stripping out a lot of that uniqueness in the language and that at some point language models are going to have to grow past the generic speak that comes out of it. And I think about even opening phrases when you're writing an article or something like that in the ever evolving world, like that is such an AI statement now. And even if you've said it yourself or delving into, even if you've said it to yourself, if you've said the word delve before chat GPT, you can't say it anymore, especially you can't write it. Lori Schafer (24:00.334) Yeah, I mean, there's no question we're in the early stages, but already, I mean, the thing I certainly believe in, I know, you know, with our company, we hire all people from the industry because technology on its own is not going to give you a natural, you know, natural voice, natural brand voice, but working to be able to use that, I keep going back to use AI. to do the mundane tasks, don't try to use AI for everything because that human touch is critical. It's absolutely critical. That's never gonna go away. And the human brain works radically differently than computer technology, AI algorithms. They're very good for repetition. They're not good for creativity. And I think even in everything I know that... we do or we touch, it's make sure that the AI piece doesn't come through so strong that it sounds unnatural. You have to have the human element. I believe in everything and I could be wrong, but I'm not a believer that you can ever take that away unless it's just a routine task that AI can do. Brent Peterson (25:23.989) Yeah, I'm going to continue to say that we produce things for other people to buy, not robots to buy. And we're selling things to people, not robots. And we're writing things for people to read, not Google. Yeah, not Google to read. So that last pass through by a human to ensure that it is readable by a human is so important. And I suppose at some point we'll be able to. Lori Schafer (25:32.142) That's right. That's right. Lori Schafer (25:37.646) Not robots, right? That's right. Brent Peterson (25:49.237) have some kind of a test on readability that's done by AI after, but even that's gonna morph into some kind of blandness because it passed it, but yeah, you know, things change. Language changes every year, and if we're not keeping, and if we are dependent on AI writing our language, then we are gonna get stuck in that year that it started. So we don't wanna be stuck in this year for the rest of our lives. Lori Schafer (26:14.606) Yeah. Brent Peterson (26:18.261) we are gonna have to have uniqueness in that language because it only knows what it knows. Lori Schafer (26:19.598) That's right. Yeah, I think the chat bots are a great example right now. You know, and there's a lot of great work being done on chat bots, but the ones I've used so far, I know right away. Get me past this. Get me to a human being. Brent Peterson (26:34.549) Yeah. Yeah. It's, yeah, and it's increasingly frustrating when you go through the chat bots questions and then it brings you in a full circle back to, is this what you'd like? Well, no, we just talked about it. Give me a tool representative and then you go through the circle again. Yeah, yeah. Agent, agent, right? So I'm almost like hitting zero a bunch of times on the phone when you had to go through a phone tree. Lori Schafer (26:52.046) Agent, agent. Lori Schafer (27:00.078) Yeah, it's a big task for any kind of e -commerce, any kind of retailer brand in the e -commerce game right now to find that balance because the technology is coming, it's here, it's not going to go away. And at the same time, that human touch, you can never underestimate that. So as I say, back to the, what is the human brain best for? What is the AI technology best for? Creativity is the human. Brent Peterson (27:32.213) If you have a piece of advice that you'd give to a retailer who has a product that they're selling and we're going into it, we're going to go into Black Friday, what would that, what would be the number one thing they should focus on this coming holiday season? Lori Schafer (27:54.318) Black Friday, right after the election. I'm thinking about the economy right now and the inflation and so forth. I mean, I'd have to say for this year, my advice would be really focus on the product value. value for the price. Just because it is, it's a rough time for a lot of consumers right now. And we have to sell the value of the product. Brent Peterson (28:24.373) Yeah, that's great advice. Brent Peterson (28:30.581) Laurie, we have a few minutes left. As we close out, I give everybody a chance to do a shameless plug about anything they'd like. What would you like to plug today? Lori Schafer (28:41.358) Shameless plug would definitely be Digital Wave Technology. I do think it's an incredible company. We are doing things no one else is doing yet. And it's because we have both that very deep, deep, deep domain expertise as well as an incredibly talented architectural and R &D team. And we marry those two together. So that would be my shameless plug. The other one would be just, I'm a big believer in intellectual curiosity and I can't emphasize enough to everybody out there, don't fall into that trap of not learning because this is now a world where you have to, you have to want to learn every day. If you're gonna look back in five years and not be left behind. Brent Peterson (29:33.909) Yeah, if people want to get in touch with you, what would be the best way? Lori Schafer (29:40.014) LinkedIn, certainly. My tag is lori, L -O -R -I -J, middle initial J, Schaefer, S -C -H -A -F -E -R. That's certainly one. LJ Schaefer on Twitter. Email Lschaefer at digitalwavetechnology .com. Any of those would be great. Brent Peterson (30:00.757) Perfect, and I'll make sure I get all those into the show notes. Laurie Schaefer, the CEO of Digital Wave Technology. It's been such a pleasure. Thank you so much for being here today. Lori Schafer (30:13.934) You too, Brent. You do look like you should be the one in Florida with that shirt on. And I should be in Minneapolis instead of Florida. All right, well, thank you so much. I really appreciate it. Everybody have a great day. Brent Peterson (30:15.157) Yeah, thanks. Yep. Thank you.