Good morning for those who are joining us from from Asia, and good evening, if you’re joining from anywhere else.
Today, we’re gonna be talking about turning traffic, right that you get into loyal customers with online personalization, right? Who am I? My name is Raj Badrinath. I’m the CMO of a company called RichRelevance(now Algonomy) as most of you guys may have known. RichRelevance(now Algonomy) and Manthan(now Algonomy) have announced a strategic coming together of forces. So very excited to be part of our new family. So I have two roles. One role is obviously running marketing. The other roles is that of a work from home husband that I’m sure many of you guys can identify with. And this is a quote, right, which I think you can identify having spent maybe just over two months and locked down. Not sure the Coronavirus is kill me when my wife is surely ready to kill me right now. So hopefully, wish me luck. Things are gonna be better at the end of this as well.
Let’s jump into you know what today we’re going to be talking about right? So this whole idea of digital transformation, right? How many years has it been? I think it’s been 10 years, since we’re talking about digital transformation. Sometimes, you know, companies pick up change on their own, sometimes change is thrust upon them. And in this case, it’s in the form of a big wrecking ball called COVID-19. Right? It’s causing people to think about how they have to relook rethink about how work gets done, how retailers have to rethink about how to reach customers. In fact, the whole bet, right, so the business transformation has been pretty much been accelerated from our point of view. Nowhere is it more prevalent, then this recent survey that Gartner did, where they looked about what was a concern for consumers a while ago, versus what their concerns are, you know, going forward in the future, it’s very clear that anything that they had, like investment value, you know, in a bull market with the markets are going very strong product shortages, price hikes, etc. All of this stuff have been subsumed by the economy. everybody’s worried about the economy, what happens now? What happens in the current what happens in the future? And there’s a few other things that are interesting, right, which is what people are also worried about is product shortages, people acting against the interest, are they going to be price hikes, etc, etc, right? So there’s a, there’s a whole shift in how consumers are thinking about the market right now and what retailers need to do, as a consequence of that. Now, despite you know, all the doom and gloom that we are reading about, I don’t know about you guys, but I’m actually sick of it, right? I don’t open a newspaper and look at how many people got infected, how many people died? It’s very sad, you know, certainly some things are happening. But the big question for us, what we have to think about is, what do we do about this, right? We are in this situation, we have to do something that’s helpful, not only to, you know, our families, we have to do something that’s helpful to our business and companies, we have to help something to drive the economy, because remember, that’s the big thing that we’re talking about. And the best way to do this is to build something, each one of us has the ability to go back to our companies and think about what we can do to create something to construct something from scratch, and do that at a very, very rapid pace, we have to do that fast. Because there’s no time, right? This thing is upon us, we have to it’s a great driver for change, and we need to go build it. But then you say, You don’t understand, you know, things don’t happen very quickly. In my company. That doesn’t happen very quickly. In my industry, there’s a lot of things that need you know, approvals for there’s a lot of things that we have to plan for, there’s a lot of things that we basically structurally are not aligned to turn on a dime. Right? So let me give you guys a few examples that you can take to your management team to say, actually, not only can you do this, there are examples across the board, where people have done fantastic things very, very quickly. Right? So this is the classic right? How do you drag an organization that doesn’t want to do what you want to do? Right, which is what we’re talking about now. An example of this is the India Mars Orbiter Mission. So you could be thinking what the hell I thought I came here for a retail specific awareness and personalization. And why are we talking about Mongolia? And, and, and things like this, right? Very simply put, right? Not only was this mission successful, do you know from end to end perspective, after all the acceleration, it just 15 months for India to launch a satellite that actually went all the way up to Mars, and they had to collaborate with NASA and everything, right? It’s working really, really well. How can you take such a complicated project and get it done in 15 months? Well, you can because there’s a little power and there’s human ingenuity behind it. Right Did you also notice that the entirety of Disneyland, the original vision of Disneyland was built by Walt Disney in just one day, over a year, in 366 days, Disneyland was completely constructed, just mind boggling to, for us to imagine right now that you can actually go from vision to execution and just 366 days just over a year. And also, did you know that the first version of the iPod between, you know Steve Jobs, you know, thinking about this, Tony Fidel, who’s as SVP, they had this conversation around? How do we build something? How do we change the music market, there’s plenty of layers, this whole thing from end to end, from design, conceptualization to manufacturing to go to market just to 290 days, right? Pretty fantastic. But if you think about a product that came and changed the world, did you also know that one of the largest field hospitals are built in Dubai 3000, bed Field Hospital was built just within one week to handle the corona situation right now. Just incredible examples of human ingenuity of how people when they come together, can create some fantastic impact. And you know what, ladies and gentlemen, what we have in front of us is to do the same kind of impact. Right? That’s us. All we trying to do is to build an omni channel customer engagement system, right? How many days will it take for us? That’s what we’re going to be talking about, hopefully, suggestions that we’re going to talk about gets you doing something very, very rapidly, very, very quickly on how to react to the situation that we’re currently in. Why is it important? Because the shift to online, right is likely to be permanent. Do not think about the shift as hey, you know, what stores are closed, everybody’s going online, the second stores open up, right, everybody’s gonna go back to stores, yes, some of that is going to happen. But some of this is going to be a permanent shift in behavior. Right? The permanent shift in behavior means that you’re going to have much more of a heavy digital presence and he ever had, right, when you look at the Asia market, most retailers have maybe 2 to 10%. 10% is actually quite mature, right? We’re not talking about the pure play ecommerce, we’re talking about retailers or classic stores, and so on, right? anywhere from 2 to 10%, sometimes even lesser is what was being done online. Today, what we’re saying is, look, this number that’s going to be you know, 10% or more suddenly starts to, you know, remain almost the same. What if people who have gotten trained into the frictionless way that you use your mobile applications to place orders, never would decide to go back to source again, right, or they do it, but they do it only 10% of the time. This means that you need to start to rethink in categories, whether it’s essentials like groceries, and household supply, all the way to you know, apparel, and footwear and furnishing, all these things are going to start to change as well as we go forward. Okay, so here are some ideas. And I’m gonna go through these ideas quite rapidly, hopefully, what this, this will do is to give you guys a framework of how we’ve thought about the problem, we work with a bunch of customers, these are all very rich, very visual customer examples that I’m sharing with in front of you guys. And hopefully, you will take away some ideas of how people reacted to, you know, online personalization. And these times, how do you take traffic that’s you currently coming in, right, and help you grow revenues quite rapidly.
If I were to summarize it, there’s only three things right that I want you to remember today, right? What do you do? Right? Three things that you do is number one, you make it timely. Time is money. Number two, you make it relevant. And I’ll tell you how you can make it relevant and rich. And number three, is you drive them to the second purchase, right? And it’s a key concept. What does it mean, don’t let the move online stay as a one off event, right, don’t squander this opportunity. Winston Churchill once said, right? You should never let a crisis go to waste, right? How do we take this and make sure that we continue to win the loyalty of our customers going forward? Why? Because it’s not just the first purchase. We want them to do the second and then the third and the fourth. Okay, let’s talk about what I mean by that.
First one, how do you make it timely?
Here’s an example. Right? One of our customers basically said that, you know, when you start to look at products that are out of stock, which is common, right, that product, you know, runs out of stock, especially more rapidly now where they are, there is a great chance that things can be lost. So typically, what we found is customers who didn’t have you know, actually had the recommendations, which was below the fold, just by moving it above the fold had a significant impact in conversion rates. Why? Because as Soon as you look for a product and you came and let’s say in this case, you’re looking for Lysol disinfecting wipes, you don’t have to scroll all the way to figure out, you know, whether you have the, you know, you had the product available or not, you could immediately see right here that look what you asked for is sold online, here are a few other products that are available that are similar to what you’re looking for. So automatically identifying what’s a in stock and be available that similar to the product you’re looking for is important, right? That helps with customer satisfaction. It’s timely, because you know that you’re valuing their time, you don’t want them to scroll through stuff, right? Number one.
Number two, not only can you do out of stock products, you can do products with similar health benefits. Right? So imagine this right? So when you look for hand sanitizers, right, so especially here in the US, it was very, very clear that, you know, people went absolutely crazy buying hand sanitizers I’m sure it’s exactly the same in every part of the world, right. Where, you know, hand sanitizers, which was really, you know, people didn’t pay a lot of attention to all of a sudden became, you know, almost like gold, right? It became so precious. And everybody ran out of everything so, so quickly. But it turns out that it’s not just hand sanitizers, right, that were needed, right, it’s very clear that it was all plain old, regular soap could work really, really well. And so what many of our retailers actually did was something very simple, right? They basically said, Look, you may not have a hand sanitizer, or do you have soap, which has similar health benefits, right, that is automatically recommended. This is always in stock. And again, the messaging is what changed dynamically for them. So that’s another way of saying, Look, how do you look for a creative way of servicing needs for out of stock? Here’s another really cool example that I really like, Okay, this comes from the wonderful world of Brazil. So about, I would say, a month ago, we had, clearly and you saw that in the chart, right? A lot of entertainment that ends up happening is all happening remotely, right. So you have, you know, we have rock stars, who obviously also have to make them make money, so they don’t live streaming concerts, with tickets that are being sold, and so on, and so forth. So one of our apparel customers did something fantastic. What they did was, they realized that you know, what, in this case, Maria Mendonca, was a was a very famous Popstar, in Brazil, what they did was they looked at the costumes, the outfits that she was wearing. And if you go to this is a customer ratio in Brazil, if you typed her name in, you’d see exactly all the dresses that she wore. Therefore, they were able to increase conversions by making it so timely. So if your Imagine your, your, your state of mind, right, so let’s say you’re a big fan of a rockstar, who’s actually performing in front of you. And basically you like what she’s wearing. Imagine if you could go someplace. And of course, they’re sponsoring and there’s marketing that was tied to it, you click on it, you come to a site, you type in her name, or the links carries the context with you, it immediately tells you the clothes she’s wearing, the accessories she’s wearing, etc. That creates a time bound contextual relevance, that is very hard to replicate. Right? So those are the kinds of things that they could do. And they use, you know, the RichRelevance(now Algonomy) product called the search product, it’s called Find in conjunction with some of other products to actually help make this happen. Here’s another one, right? Remember, the biggest thing that is a source of frustration sometimes is, we are all creatures of habit, right? It’s very rare that you go in, and you want to buy new things, especially for essentials, like groceries, right? Typically, I go in, I know, I’m buying milk, I’m buying eggs, I’m buying bread, etc, right? These are some basic things that you end up buying all the time. Imagine what happens to when you go to an E commerce site, right? Today, it’s almost as if the website has no memory. Right? You go in and whatever you bought, you have to redo the whole thing all over again. Right? What if there was a way where the engine knew not only what you bought frequently, it also knew the frequency of when you bought it, so meaning, if I know that, you know my wife sends me sometimes to stores to say buy makeup wipes makeup wipes, she uses them roughly like once every six weeks or eight weeks. Okay? What if the engine could detect that my maker wipes is not like a weekly purchase? It’s not like every week, I’m gonna go buy something, but it’s automatically able to detect and tell me that, you know, around the fifth week or whatever, when I’m more likely to make a purchase, I can show you those kinds of things automatically. This is the power of AI, this is a replenishment algorithms. So the whole idea is help make you know commodity purchases by automatically determining the intervals and actually start to show that without you having to do anything, so this is even more important now, when you start to think about how much things are changing how much the purchase behavior, it’s also changing right now that, you know, people are stocking up on toilet paper people are stocking up on, you know, all kinds of you know what you saw earlier, right? X soaps and things like that. What do you need? What is the frequency of purchase without having to sign up for a subscription plan? Right? Remember, subscription plans are passing away, but there is something called a peak subscription plan, right? How many subscriptions are you going to have as a consumer? Not a lot? Right? You try? That was a great trend. But it’s all it’s all falling apart now. Right? So how can you make it easier? How can you make it more timely, without having to have a subscription plan? Right. And not only that, the last point over here is suggest other things substitutions. If something is out of stock, that’s another cool thing that we could do as well. Here’s another way, which is more and more timely. And this is more from an implementation perspective. Many times, making a change, as simple as it could be, is it’s hard because you know, there’s an IT process, there’s an approval, and so on and so forth. So when I talk to, you know, marketers around the world, right, what they tell me is, look, I wish we could do something ourselves without having to rely on a developer organization without having to put all those things in place. And of course, I needed to be done in a way that’s kind of protected. So I don’t screw it up. And I don’t destroy the website and the process, right. So one of the things that we found the customers doing is making changes in real time with this tool called Experience Designer. What is it what it is, is very simple. It’s a drag and drop interface, where people can show in all the latest images, promotions, etc. And just by creating this, right without having to use any of the technologies, you can push out the changes, and these things are live right away. Imagine the time to market advantage that you’re going to get if you’re able to say look, I have real time reporting that’s coming in, I see my website trends that says these things are actually becoming more important. I’m going to drop a quick placement over here that automatically caters to not only the new messaging, but also the new products and attributes that I care for that I’m promotion that are promoting for personalization right now, all of this is possible with Experience Designer. Okay. Now, let’s go to the next big bucket. Right, which is, not only did we make it timely, last he saw, there’s different things that you can show. This is even more critical in one sense, right is how do I make it relevant? Right, let’s look at some examples of how you can make some things very, very relevant. This is a is a feature called affinity bias recommendation. That’s a mouthful, right? So what is affinity bias recommendation? The affinity is very, very simple, right? Affinity is what you like. So if you are a shopper and you go to website, you will exhibit an affinity towards a certain thing, right? That could be a brand, maybe you have a favorite brand. Maybe you like Nike sneakers, maybe Nike is your favorite brand. Or you like a particular category, right?
You’re there to buy, you know, shoes, just keep it the same analogy. You’re shopping for shoes in that session as primarily, right. So that’s your category that you care about. You don’t care about the rest of them, you care mostly about shoes, especially for the session that you’re currently in. Or it could be any other product attribute that you care for. Imagine if right, if let’s say you’re a woman, right, and you go to a website, in this case, you go to REI, and you’re only looking for shoes, right? The engine now can tell that you know what it because you’re a woman you gravitated towards that it makes every other product that is not relevant to you, meaning it’s for men’s, it’s for teens, it’s for kids, all of that stuff are removed from you. And only thing that you see are women’s products. So you could do that by increasing the weight age that the Indian actually gives towards a particular affinity that you have. And what we found is an increased attributable stays by 38%. Think about that for a second. Just by making it relevant. You have 38% more revenues than what you had before. Just absolutely incredible. And all this in real time. No code required only a configuration. Right. Here’s another one, about cross selling. Think about in the grocery example I used earlier, right? You’re searching for, in this particular case, a pizza base right. Now your two ways of approaching the problem. You can say oh, you know what? This guy search for a pizza base. Let me show him all the pizza spaces that we have. That’s great. That’s okay, that’s expected. But what if you’re able to divine the intent of what they’re trying to do? Are they trying to throw a pizza party. Which means not only do you need the pizza base, you need all the things you need to put on the pizza available to you. I don’t know about you guys, every time I have a pizza, I love to have a coke with me, right? I love to have a can of soda next to me. So that’s something that goes along with it. Because this is what the engine has determined over time. Imagine if you could do advanced merchandising that says I know exactly what items travelled together, your merchandising teams can work with the AI to then mix the affinity of what the user is doing the intent of what the user wants to buy, in real time and create ko purchase. bundles. Right? So essentially, what’s happening over here, in this case, I misspoke. It’s not just bundles, it’s about showing products that travelled together so that way, not only are you improving the size of the basket, right? Clearly, you’re not just buying one thing you’re giving more. But also you’re delighting the customer, because they’re saying that, wow, this is like a real salesperson who’s helping me digitally. Remember this still purchasing? You know, it’s a very cold experience. It’s not like walking into a store, right? Because there’s nobody there to help you. And even if you want to ask someone, who are you going to ask right? Who’s going to have context awareness of what you’re trying to do? This is what personalization truly does is brings relevance, but trying to understand what you’re trying to do in real time with the right kind of messaging. Let’s go to the next one. How do you then curate category and search pages? Right, which is in context of what you’re trying to do? It’s very important, right? The problem you’re trying to solve is that look, most retailers have, you know, hundreds of 1000s of SKUs. Right?
It’s not a it’s not a small problem to solve, because you’re not, you know, unlike physical stores where you can walk through aisles, and so on and so forth. Typically, what you end up seeing is there’s a screen in front of you, it gets worse when you have a mobile device. But the whole idea is when I search for something for a catalogue, product discovery is a real challenge. Right? How can you then make sure that when you search for something, how can you resort these products based on the shoppers behavior. So it looks at what I’m doing right now what my interest is? And think about it this way, right? Imagine if I’m vegetarian, okay? Or I’m gluten free? These are attributes that you know about me, right? What if you’re able to show only gluten free products? Because I declared that once right, in my preferences? How do you exclude those products. And imagine there’s another scenario where you know, the buyer team, the procurement team actually bought an excess inventory, and you want to intersperse excess inventory because you want to move that product, but you want to do it in a way that customers actually look at this as discovery of new things, as opposed to complete the whole set of irrelevant results. You cannot do these things, guys with rules, right? Most people think about this whole thing as hey, I’m just gonna put a rule and merchandising rule, a marketing rule that’s going to help everything and promoted rules or static rules tend to break, what you need is contextual awareness. This is where machine learning and AI actually comes in really well. Right? It layers it. So it’s not just every product is a rule. It’s like one of the products is a rule, but in context of what we’re trying to do so that way, it’s highly, highly relevant. Okay. Here’s another one, which I which I misspoke earlier, right, which is the idea of creating bundles that’s automatically curated. So here’s an example of somebody who’s trying to buy a Mac, right? They’re looking at a MacBook Pro 13 inches, right? 256, GB, SSD, etc. What if you’re able to automatically create and you’ve seen this in Amazon, and some of these guys do that? Right? So what if you’re able to create a whole bundle of what you know goes along with it with one very important attribute and most people miss is compatibility. I don’t know the number of times I’ve spoken to customers that says, Hey, we are in the consumer electronics space, right? We’re looking at this MAC, many times that attribution that comes in right it shows you different products that are actually supposed to be bundled with it turns out that none of them work with a Mac, they’re all windows parts, right? Imagine what happens to the customer satisfaction when they buy something right a part over here that looks like it’s compatible, the add to cart to buy it and immediately they’re frustrated, it doesn’t work, you know the things so poorly the return it is a whole process, they call the call center, there’s a return process, it eats out your margins, the whole thing right? How do you do that automatically? And how do you do that in a much more relevant manner? Those are the kinds of things that helps right even in this case right now, in improving relevance. Not only this, by the way, these things can be extended to the call center as well. So if you have a call center that does out about you know selling, then these things can be easily be exposed to call center as well. So they have the, you know what the customer did website on the website before they come to the call center. Here’s another example, right this is called personalizing for cold start, what is cold start? Cold Start is a simple way of saying, If you come to my website for the first time, and I know nothing about you, I’m starting cold, right? I’m trying to figure out what do I show you that actually interests you? And this is a hard problem to solve, right? We have solved the problems in multiple ways. But one way of solving the problem is to say, Do I know where you came from? Do I have a referrer? Right? Do I know that you came from a social media site? For example? You came from Twitter or Instagram? Did you come from a campaign that we were running as part of marketing? Right, that got you back into this? Now, once you came in, I know your geolocation. So I know, roughly, you know that you came from, you know, Bangalore, and I’m hearing that it’s very, very hot in Bangalore. So clearly, you’re not going to be buying winter coats or sweaters, right? I know what device you’re coming on. So that way I can, I can tailor it based on the form factor.
I know, you know, what you’re currently doing in the session, and so on and so forth. Right? So what the system then does is it starts with showing you what’s selling, right now at this instance, what are the top selling products for your region, closest to your store, from where you are, right? Those things come up. So right here, not only that, it starts to show you what’s recommended based on your current session, what categories you need to look at, and so on, and so forth. So the ability to control this is actually very, very flexible in terms of what we have. The main point here being is that you need to start to improve relevancy, even when you don’t have any digital signals, right from the very beginning, right, because that’s when this whole thing is actually going to work. Right. Let’s keep going right on a couple of other things. The third bit and the final bit that we talked about, right, we talked about how to make it timely, right? We talked about how to make it relevant some ideas on all of this. And now we’re going to talk about how do we make online and a habit and drive customers to make a second purchase? So it’s not just the first one? How do we make it to a second purchase? Right? This is a really cool feature guys that we did about like to say about a year or so ago, which completely changes the way you think about personalization, the way it’s been done. And I don’t think anyone in the market actually does this. Let me explain what I mean by that. Imagine a scenario where you have you know, it’s, it’s a Christmas sale or a Diwali sale, or eat sale, right, something’s happening. It’s a big event, you got a whole bunch of new products that came in new seasonal products that just came in. And you need to make sure that these products are, you know, being surfaced, there’s they’re being discovered, and they’re being matched to what the shopper is kind of looking for. Now, the way to do this is typically marketers will end up writing a promotion and saying, hey, you know what, we need to make sure that these things that’s a very prominent billing, for everybody sees it, and so on and so forth. Well, it’s not very personalized, you can say, hey, here’s a Diwali collection, click on it, everybody gets 50% off, while you’re leaving money on the table. For those people who actually don’t care, they get irrelevant. You know, you can do some A B testing of content and so on. But the real interesting way to do this is to make sure that you are matching it to the individual customer who’s coming in. I may be specifically looking for something that other minorities, the other person may not. So why am I seeing the same thing, right? But how do I do that without having to force everyone down the same path? Well, we do that through NLP
What is NLP? NLP basically does something very cool which is it does not wait for behavioral data for personalization. When you look at the personalization market today, every tool that we have in the market looks at person to look at looks at behavioral data, excuse me, meaning it looks at what you clicked on what you looked at what you purchase, etc, as a way to make experiences more relevant, which is nothing wrong with that, which is still the strongest way of doing it. But now what we have is we brought in a new deep learning feature, which is, you know, a fairly advanced AI, with neural networks, and so on and so forth, to create associations between text descriptions of products, and the buyer shopping behavior. So we’re now able to create associations between what you’re looking to do so here’s an example of a company called Blue tomato. It’s an Austrian sports provider, right? So what blue tomato does is when they’re turned on NLP, it’s got two kinds, by the way, so when you have behavioral data, behavioral data means that you know, people have actually purchased this product before in the past, somebody looked at in the past, so there’s a good correlation that the engine can do to say, okay, look, you bought skis in the past, maybe skis is what you’re going to be interested in, right? But what happens now is without behavioral data, new products, obviously have no behavioral data, why there’s a new there’s something called longtail products, longtail products are things that are your in your inventory that haven’t been bought. That’s the reason why they’re a longtail because you know, they don’t see much traffic, but you still have inventory costs for them. But sometimes what ends up happening is, that is exactly what you’re looking for. So if you buy a skateboard, and you have a longtail product of skateboard replacement wheels, for example, right, this engine can automatically make it relevant and actually surface that without you having to write a rule for it without you having to write a special promotion for it, and so on. And it does that automatically and learns over time. So what we found is not only does the click through rate goes up by three points, percent, right? But longtail product RP live went by two by 2%. We have another customer where the cost, the revenue went by 100%. So there were about $200 is what they made per visit, it went up to $400, simply because we use NLP. So NLP guys is a is absolutely a game changer in terms of how you think about personalization with non behavioral data. And I’m happy to talk to you guys afterwards as well if this is of interest to you. Here’s another way right of how do you drive them to the second purchase? Remember, it’s not just about the first transaction, but it’s about becoming the trusted adviser online to educate and engage them. Here’s an example from Brazil again, there’s a company called epoca Cosmetic goes, they’re one of the largest cosmetic companies in Brazil. They’re like the Sephora of Brazil, if you will. So in this case, if you search for the word beard, right. And if you’re a guy, and I’m a guy, I have a beard. So I That’s why I picked this example. You can see there’s a bunch of things that you get right, you get some shaving cream, you get like, you know, beard freshener, and whatever I didn’t even know there were so many beard products, by the way. So you get a bunch of things over here, right? But not only do you get that you also get very helpful videos that tells you how to maintain a beard that tells you, you know, what’s the right way of doing things? How do you use the products? How do you apply the product, etc? How do you trim your beard, especially important in the corona lockdown days, right? We’re all getting our own hair and trimming their own beards at this point. So all of these things can happen when you combine your search to not just have a product search, but also have content along with that. And that’s how you create engagement. That’s how you create education. Right? And remember, customers, once they find out that not only are you selling them a product, you’re selling them an experience, guess what, that’s where the loyalty comes from. Right? That’s when they go from Hey, not only is this guy giving me a list, right or pages and pages of products, they’re actually spending the time to teach me on how to use it. And that’s where you know, it’s going to be make the difference between you and your competitors. Here’s another one. How do you personalize the homepage content for specifically for the brand, especially when space is limited, and so on and so forth. So there’s a lot of text over here. I’ll give you a simple example. Look at the bottom right over here, right? These are some very famous brands that most of you may be familiar with Electrolux, which does white goods, PS is for PlayStation, which is by My favorite Dell and of course LG right now, PlayStation, let’s use that as an example. So many times can you talk to people who are gamers, right? So a big confession. I’m a gamer, too. I love gaming. Sorry, don’t kill me. But I’m a PS gamer, right? I’m not an Xbox gamer. I’m FPS gamer. So when you say you are a gamer, it’s very important that you don’t start to merchandise people to the wrong console. So if I’m a PS gamer, and he showed me 100, Xbox games, guess what you just made my experience highly, highly irrelevant, right? Because none of the Xbox games will actually work on a ps4 or a PS five, that’s going to come right? It’s not compatible. So all this is saying is that if I show a very strong signal that I’m interested in a PS, change everything. So that way, in a mobile device in a limited space, you’re re emphasizing re ranking and arranging the brand, based on what I’ve told you. So it replaces all my default brands, it goes away, if I’m an apple customer, don’t show me Windows stuff. If I’m a gaming customer, don’t show me non gaming stuff. So the whole point here is that you can train and tweak the engine to make it super relevant. And super quick for someone. And this, again, is a way to say I care about your preferences, and therefore I’m going to come back to the site again, because they seem to know what I want. And this is a great partner and customer of ours based in the Nordics. Here’s another way. There are many kinds of data that you ended up capturing, right in the online session. Some of them are implicit, like the affinities that we talked about earlier, where the engine tries to figure out what is it that you really like without you have to tell us? Some of them are explicit, explicit is things like preferences, guess what it’ll ask you? What is your eye color, especially for women who are looking for makeup? Right? It’s very important that you look at why color and so on. This is L’Oreal, right? Which, which is what they do on their site there on the mobile device. If you specify that my color is green, do you think the next time you go shopping, your eye color would have changed? Of course not. Right? Now, yes, you can argue that I’ll be shopping for someone else. And maybe it can happen. And there’s a way to do that as well, for the most part, right? You are a loyal customer, you want to come back, you will have a set of preferences that are unchanging, it asks you things about what’s your skin type, like, right? You have oily skin, they have dry skin, it’s very rare that you go from dry skin to oily skin the next visit, right? How do you make sure that every single time the engine is able to show only relevant results to you by matching your preferences with the product catalogue, and making that connection in a seamless manner? Right. That’s what this means.
Here’s another one. This is another cool thing that we did, but a year and a half ago, which is automatic discovery of micro segmentations. Right? What does that mean? Think about your traffic, right? You got to have a whole bunch of people coming in. Some people are coming shopping online for the first time, you know, your grandmas and grandpas will go buying things on mobile because you know what the hell nothing else is working. If they can, they can go to stores. Now, you’re now getting a wealth of data behavioral data about all of these guys who are coming online for the first time. Now, wouldn’t it be great if the system could automatically bucket these guys into different segments? And this happens in real time? What do I mean by that? Here’s a quick example. Among all the people who came in these are audience right? Right across 43,000 sessions or so 14.6% is what the audience is made up of. And what happened was that the average order value for these guys, I’m just using this as an example was 47%. Lesser, right, but the conversion rate was 88%. More, what does that mean? It means that these guys buy a lot for the buy cheap products. So these guys are bargain hunters. Right? You can easily create a segment over here and create a promotion against those segments as well. Now the inverse is true as well, meaning you have you know, very, very high AoE. That means, you know, they buy expensive products, but the conversion rate is very low, which means they are perfect crossover candidates, right? You need to give them something that actually improves the conversion rate as well. You can create a segment or here or you can discard it. The engine constantly is trying to figure out how do I give you opportunities to increase revenue? And how do I do it in a very seamless way where you can tailor your marketing campaigns completely based on this behavior. And you can not only take the marketing campaigns online, you can take them offline, you can send them emails afterwards, right? You can send a print mail campaign afterwards, whatever you that you end up doing right there. Right. But everything is driven from how customers behave. And you’re doing that dynamically. Okay, so what is it that you shouldn’t do? Right? And again, there’s not a lot of what we should do, there’s really three simple things, right? First thing you shouldn’t do is, wait, why? Because as I said earlier, this is the right time to implement new things in place. Right?
Number two, you have to experiment. Right? This is the time where use your engine to try out new things, AB test them, right? Do your MBTs get these things out there and have an engine that lets you experiment very rapidly, and is able to take a commerce point of view, because ultimately, it’s about driving revenue. Look, when the stores are short, and digital becomes a primary revenue driver, this is a unique position that nobody in the digital space has gotten before. This is, you know, something that’s never happened, because it’s unprecedented, right? Why would you not then show the power of experimentation, right? With a tool? This is the time that you can do, right? And imagine that this is the new normal, right? We’re never going to go back to exactly how it used to be, this tool is going to become more and more prominent, right? So don’t wait to think about, you know, there’s going to be normalcy, we’re going to go back to how it was. And you know, maybe I can I can I can think about these things. Right? That’s my summary. Right? So this is I’d love to sort of open it up for questions. Let me just bring it up into the three things that we talked about today, right on what to do. And three things on what not to do. What you do is, number one, make it timely. Number two, make it relevant. Number three, think always about how we drive to the second purchase, not just the first one, right? Everybody can do the first one can be even be an accident, right? But how do you get them to the second purchase is important. And then what not to do is don’t wait, experiment now. Prepare for the new normal. Again, to remind you what Churchill said, Don’t let a crisis go to waste. And when we have fantastic human ingenuity projects that we did before, which can be done. If you can build Disneyland within a year. I’m sure you guys can build something for personalization within 30 days, right, if not lesser. Right. Thank you very much for that. Let me just open it up to questions if there’s any happy to answer questions for you guys.
Perfect. So there was one question here that says are you going to talk anything about the virtual stores? I’m, I’m not quite sure what you mean, virtual stores? Maybe you want to clarify that I was. I think we’re going to I think I spoke about some of them. But let’s, let’s come back to that if they need clarification, I’m happy to answer Okay. Next question, Is it is it all is really time consuming to set up a marketing and merchandising? Absolutely not, right? I mean, the whole idea for marketing and merchandising is the idea of how AI and algorithms work with the marketing teams in their ingenuity with their experience with their curation skills, and make it work together. let the machine do the work in terms of the heavy lifting, and guide the final output through marketing and merchandising roles. That’s how we make them to work together. And there’s another question here, which is how do retailers measure breadth and depth of online personalization? visa vie, so many other options available? Yeah, that’s it. That’s great questions, right? So look, the idea of how you look at online personalization. Right? Is isn’t multiple layers, right? You can go from, you know, very simple rules based mechanisms that exist that all say we are, you know, we are personalization tools. You can go with startups right now, it says there’s plenty of startups that says, hey, we can do personalization. And what we have, remember the number one way of looking at the depths of personalization, which is what I’m answering now, the breadth is the quality of personalization is defined by the models, right models is there’s no, you know, it’s all data science, right? And data’s data science and the quality of models is defined by the traffic that these models get exposed to. So they become better and better in terms of the predictive abilities. Right? And so, if you take a model that has not been refined, has not been, you know, put to the test of time across multiple retailers around the world. Guess what the model still desires a lot to be trained. Right? And this lack of training is the difference between a decision that’s made by a three year old child and a 30 year old man. Look, decisions can be made by both Right, a three year old makes decisions, right? often not very good ones. But a 30 year old man can make decisions as well, right? So the ability to quantify the difference between maturity of data science and AI is a huge distinction factor, right, which retailers need to evaluate. Now, in terms of breadth, look at the amount of use cases that we can do, right, whether it’s brand awareness use cases with respect to engagement, whether it’s conversion related use cases, whether it’s margin management, use cases and optimizing for that, or replenishment use cases, all of these things, we have about 150 Plus strategies out of the box that are every conceivable, you know, use case that you can think of is actually available out of the box, right? That makes it hugely different from everybody who’s like, Hey, I have one or two algorithms, and you have to write a lot of developer activity to actually get that going. So that’s kind of the point of how you think about depth and breadth over here. Of course, there’s a lot more detail, but I think this is going to probably suffice for now. Any other questions? Guys, really good. Thank you for engaging with me.
Okay, so if there aren’t any, what I will do right now is we will end this, I will give you back 10 minutes of your time. And thank you very much for participating today. I hope you found this useful. And I’m sure we’ll follow up with you guys with recordings and things like that. Oh, wait, there’s one more before I dropped off. Okay. Here’s one last question. Mobile app offers lesser real estate, what’s your view on perception mobile app. Now, look, mobile application is the number one reason why real time personalization should absolutely be implemented tomorrow, right? Because the dot only is a form factor, something that is, you know, constraining. So what you could show on a desktop, think about, you could show maybe 20 products, right? That may be good recommendations for what the customer is trying to do. But on the mobile app, you can show them three, you can show them maybe six, right without the you know, the resolution becoming so small that you can read what is being shown anymore. So it’s even more critical to bring in what we call the concept of hyper personalization. What is hyper personalization is really three things. Okay, number one, it needs to be tailored to the individual, meaning I know the individual’s affinities I know their preferences and other intent, right. Number two, it has to be driven through AI, I now can treat you as a segment of one, because of the compute power have an AI right with Bheeshma machine learning models, I could not have defined an entire journey, just for one individual. It doesn’t make sense, right? But now the machine can do it, we should leverage it. Number three, it needs to happen in real time. Right? So with these three things are achieved. It’s called hyper personalization. And this is where mobile becomes the testing ground for it. Because what you’re now doing is you’re making a bet that in that small real estate space, I’m going to show you the most accurate most relevant products, given everything I know on the background. Right. So that’s something that’s pretty critical. Okay, any other questions, guys? Okay, awesome. All right. Thank you very much, guys for attending. Appreciate your time. Appreciate your attention. And I wish you a wonderful day. All right, we’ll talk to you guys soon. Take care. Bye.
