Show notes
Everyone is excited about A.I. and the idea of using technology to improve business. But that excitement has also led to confusion. There are many definitions and applications of A.I., and few have been able to truly optimize their A.I. strategy. That’s where Ashwin Mittal comes in. Ashwin is the CEO of Course5, a transformative intelligence company helping companies improve their businesses using technology like A.I. and machine learning. On this episode of Up Next in Commerce, Ashwin explains where and how people have gone wrong with A.I. in the past, and the steps an Ecommerce company needs to take in order to be able to get the most out of A.I. It all starts with understanding and controlling your data, and includes retraining your employees to rely less on their guts, and more on the analysis A.I. provides. He reveals how to do exactly that on this podcast. 3 Takeaways: It’s important to build a structured data foundation from the start so you can move to a place of making decisions with data You must teach your employees how to interpret and analyze data and make sure they are comfortable using the data to make decisions. Individuals must be taught to lead with data and then let their gut take them the final 20% of the way toward a business decision Understanding the entire customer journey and where there are points of failure will help you create a strategy on optimizing your data and building an Ecommerce experience that is successful from start to finish For an in-depth look at this episode, check out the full transcript below. Quotes have been edited for clarity and length. --- Up Next in Commerce is brought to you by Salesforce Commerce Cloud. Respond quickly to changing customer needs with flexible Ecommerce connected to marketing, sales, and service. Deliver intelligent commerce experiences your customers can trust, across every channel. Together, we’re ready for what’s next in commerce. Learn more at salesforce.com/commerce --- Transcript: Stephanie: Hey everyone, and welcome back to Up Next in Commerce. This is your host Stephanie Postles, and today we're joined by Ashwin Mittal, the CEO of Course5 Intelligence. Ashwin, thanks for coming on the show. Ashwin: Thanks for having me, Stephanie. It's a pleasure to be here. Stephanie: Yeah, I'm really excited to have you on today because throughout a lot of our previous interviews that we've had so far, everyone has mentioned AI in some form or fashion, and when I saw that you were coming on the show and we were going to have a deep dive conversation into AI, I was very excited. Ashwin: It's a deep subject to talk about, so yeah, it should be fun, hopefully. Stephanie: It will be. So, what brought you into the world of AI? What got you excited about that and building a company around AI and analytics and all that? Ashwin: So, we've always been in the business of delivering insights for sales and marketing to customers from data and information. But about five or six years back, I realized that we're in this perhaps future technology wave of transformation because of the onset of artificial intelligence technology, and from a big picture perspective, if you look back the last 30 years, 40 years, we've perhaps seen two or three waves of substantial change value creation through technology. We've seen the PC wave, and we've seen the internet wave. You can say, okay, maybe there's a separate mobile internet and social media wave, but it's part of that. Both of these have completely changed our personal and professional lives, and one was built on the other. Because we had PCs, so the internet was [inaudible 00:02:03]. Ashwin: So, my belief is that with AI it's going to be pretty much the same. It will take 15-20 years to play out, but its impact is going to be as deep and profound as the internet. And it's going to be built on top of the fact that we all have computers, that we're all connected, all devices are connected, and for us being in the data business of driving insights from data, we're really in the eye of the storm. Because the root of AI comes from data science. So, on the one hand while this presents a challenge to our business, and possibilities of us getting disrupted, but at the same time, it presents even larger opportunities. Ashwin: So, at that time, I decided that we needed to really double down on the AI wave, and started to completely reorient the business and the company towards riding this wave. So, that's really how [inaudible 00:03:12]. Stephanie: That's great. So, how would you explain Course5 Intelligence? What do you all do? What kind of clients do you help? What problems do they come to you with? Ashwin: So, we help our clients make the best decisions from data and information for sales marketing customer-related problems. And within that, one of our largest focus areas is digital and Ecommerce. We typically work with large corporates, Fortune 500, Fortune 1,000 companies, and in some cases, we might be working with the Ecommerce group, their digital group, with the CMO's office, sometimes with IT or some combination of all of these. Stephanie: Very cool, and you guys have over 1,000 people that you employ, right? Ashwin: Yes, we have 1,200 people. Stephanie: Wow, that's crazy. When did you all start, and how long did it take to get to that 1,000-person mark, and what was the change like within your company? That's a lot of people to manage. Ashwin: Sure, sure. Yes. So, we've been in business for, what? 16, 17 years, and we've also changed and evolved along the way. The way you run a company that is 50 people doesn't really work when you're running a company that's 200 people, and you need to run a company differently when it's 500 people, and certainly when it's 1,000. So, we've also had to learn, change, evolve along the way. How we run the company, what kind of talent we bring in. How we set the organization structure, and today we're pretty much a global business. We're present in seven, eight different countries, have customers all over the world, so we're sort of a micro multinational on our own. Stephanie: That's great. Is there any struggles that you face when it comes to working with your teams around the world that you had to learn along the way? Ashwin: Sure. Yes, yes. No, absolutely. Absolutely. So, there's one challenge, which is a more difficult challenge, which is culture. There's a second challenge, which is also important, but perhaps not that difficult, which is regulation. Culture is a really tough one. It's not as tough today as it used to be maybe seven, eight years back. Today because of just the globalization and easy access to information everywhere, people are a lot more aware of different cultures, different people, and a lot more sensitive and things like that. Ashwin: But in our early years when we were globalizing ourselves, we had to learn about different approaches, different cultures. We had to be very open minded, and I think we managed okay. We're still learning, it's important to be sensitive. Culture is an important issue. In terms of regulation, that's somewhat easier. You just have to follow certain processes, protocols and make sure that they're completely compliant in every job that you're offering. And especially in a business like ours where you're dealing with data and information, and there's a lot of regulation around that, around data privacy and sensitivity around that and things like that. So, we need to just ensure that we are compliant in every geography that we operate. Stephanie: That makes sense. So, everyone talks about AI. Like I mentioned, a lot of people mentioned in our previous interviews, and oftentimes I think people if they're using AI or they're referring to something like it's AI and it's actually not. So, how would you describe AI? And specifically, how does AI help Ecommerce, or could it help Ecommerce if people aren't already looking into it? Ashwin: So, first point you made is spot on, right? What is AI? Stephanie: Yeah. Ashwin: So, this really means different things to different people, and honestly, there is no one perfect watertight definition. So, AI is essentially an umbrella term that covers a bunch of different technologies, which are all meant to convey computers getting intelligence, which is not typically expected from computers, and which is more human-like. Which is the common sense definition, which we would all accept. But then what falls in that bucket, and what doesn't fall in that bucket is different for different people, and I'll give you a simple example. Take a very old technology, something like OCR, optical character recognition. Nobody would call that AI today, but when it first came out, maybe some people thought it was actually computers becoming intelligent, and it was kind of AI. Ashwin: So, the definition is also keeping on evolving as our expectations from technology keep changing. But I would say that amongst the different technologies that encompass AI, the core one, the most important one, at least to me, is what you call learning or machine learning. And machine learning is essentially the ability of computers to learn, and that is really a game-changing technology. And then that supports all the other technologies that are typically and the umbrella term of AI. Where a computer can run a certain process or program and get better at it every time it runs, which is what humans are good at doing. Ashwin: So, that for me is one of the most important technologies. And sorry, you had a second question you asked about AI and Ecommerce? Stephanie: Yes. Yeah, how it impacts Ecommerce right now. Do you think people are using AI efficiently or what do you think the future could look like if they do start implementing this in a better way? Ashwin: So, it's already here. It is being used extensively, more by some companies, less so by some. Companies like Amazon are, of course, using it in a very, very mature and sophisticated manner. But various types of companies are using it, many of our clients are using it, and its footprint is increasing. It's used in automating various tasks, in virtualization, in campaigns, even at the back end in supply chain, inventory management. We all know that chat bots are used for customer support in A.I. In Ecommerce. Physical robots are used in warehouses, which also could have AI technology. If it helps, I can give you a couple of examples- Stephanie: Yeah, I think exactly. That'd be great. Ashwin: ... of a couple of our platforms- Stephanie: Yes, please. Ashwin: ... which truly you would consider as AI. So, we have this one platform called Compete. Now what Compete does is it enables our customers, our clients to gather competitive intelligence in the market, and respond in quick time. So, it has technologies and a bunch of bots that go out and search all competitive websites of a brand to collect and synthesize all the information on what the competitors are doing in terms of the five Ps. So, you traditionally have four Ps, product, placement, promotion and price, and we have a fifth, people reviews. Ashwin: And so, this keeps collecting this information in rapid time, and on different product SKUs and different combinations, how competitors are doing, what they're placing, how they're pricing. And then this is updated in a dashboard along with analytics to help the Ecommerce players monitor competitive moves and respond quickly so they can optimize their revenue, their profits. This platform's also used extensively during these major selling seasons, like Black Friday, Cyber Monday, and brands expect to make substantial sales in a very short period of time. Ashwin: So, they have to be very responsive to what the competitors are doing and what's happening in the market. So, this is one example, one platform that we have. Another one, which we call Adomate, this is really taking it to the next level where we're using AI to optimize creativity? Stephanie: What? Ashwin: Creativity, exactly. So, AI is not getting creative, just to be clear, and I'll explain to you how it works. Stephanie: It's not going to take over the world, do all the artwork in the future, write all the books. Ashwin: No, no, no. Well, there's some of that talk going on, but I think that you'll always want to read the book written by the author, you'll always want to know about the author. You'll want to see the artwork, which you want to know the painter's story. While those things might happen, but I don't think that's going to... Robots can play sports, right? But you want to watch humans play a football match. You don't want to watch robots play, they might play better than the humans. So, it's the same thing with art. Ashwin: But in terms of this platform called Adomate, so what we're doing is we're actually optimizing the process of content creation for either campaigns or advertising. So, when companies typically do campaigns online for Ecommerce, they will create, let's say, 100 different creatives, 100 different images, and put them into a system. And then the system will do live optimization depending on who's responding to what. Change which creative is being shown to whom. But you might have missed something fundamental, right? You don't know why something is working, why something is not working. Maybe you missed something in those creatives. Maybe there's some combination of both that you didn't know about. Ashwin: So, what our AI system does is using computer vision it actually reads every creative. So, it actually looks at the creative and then distills it into structured data. Who are the protagonists in the creative or what's their ethnicity? What's their emotion? What the background colors? Is the brand shown? Is the logo shown? Where is it shown? If it's a video, then it will break it into frames and read each frame. And so, you've got now structured data against each creative, and then you have the data of how people responded to that. Who clicked, who bought which segment, and you can combine all of that and actually then get prescriptive. Ashwin: So, depending on what kind of campaign you want to run, what segment you want to target, you can actually advise, "Okay, use a dog in this or use a Latino woman or bring the brand in or use red color instead of purple or whatever." Stephanie: Wow, that's interesting. Are a lot of companies using that today where they're actually personalizing the image for the product based on who comes in using AI behind the scenes to come up with that in real time? Ashwin: So, this is an early stage platform. It doesn't change the image in real time, though that's also possible. And my AI head tells me that we should be doing that next. But it doesn't change the image right now, what it does is it prescribes. So, for the next campaign, when you want to create something, then you would enter into the system what you're trying to achieve and the system might advise you what to do. Or if you have a creative you can submit it to the system and it will give you a score of how successful it's likely to be. Ashwin: But even that is an early stage platform. This is one of our exciting newer products, and we have maybe three or four customers using it right now, but it's certainly something that we're betting on. Stephanie: Got it. So, when these customers come to you, one thing I think about is when you have a problem. Back in my old days at prior companies, the teams I worked with who were focused on machine learning always t…
Full show notes at the publisher