The Power of Data and Personalization
Share it on
How customer centric marketing is supercharged with data and personalization
A session by Jim Kelly – Global Principal Solutions Consultant, Algonomy
eCommerce Expo, ExCeL London
Morning everyone, and welcome to the show. Uh, this is the first session of the day. We’re gonna be talking about how customer-centric marketing, uh, can be supercharged with data. Um, I’ve got you for, I think it’s three hours. Um, no, I haven’t, I’ve got you for 20 minutes. So, uh, we’re going to be talking about how, uh, the customer experience isn’t, isn’t really what it should be, and how we can accelerate and change the entire process to make it what it should be. So the impact of personalized experiences are well known. They’ve extremely well documented. Uh, I’ve got some metrics here, which you can Google. They’re all over the place. Um, 80% of consumers are more likely to buy from brands that provide a tailored experience. There’s evidence, right? Any of you doing, uh, online recommendations, that’s a personalized experience. So if you are getting an uplift on your basket size or your basket volume, if you are generating more revenue, if you get more loyalty, that’s a personalized experience. Each of these metrics, uh, really tell the story. But, um, the trends that we’re seeing in the marketplace from 2021 now moving into 2022 and beyond is is that, uh, although promotional spend is still number one on how you spend your marketing budget, second down is the huge investment that everyone’s making across retail and e-commerce for brand storytelling and, um, environmental personalization. Now, brand storytelling is all about how we engage with the consumer and try to tell them that ethically and morally, our brands are great. We source products from, uh, extremely great nations. Uh, we use fair trade, et cetera, et cetera. You are trying to create a personal conversation with a person, as you would do in a store a hundred years ago.
Guess where I got this bread from? My flour came from here. This is my cotton. It actually came from a fantastic place, but when we personalize, we personalize in silos by channel. And what’s missing is, is, is that although we are, uh, doubling down on personalized experiences, when you try to connect to a consumer, try and think as a consumer, it is tough to get right, but your challenges may not be what you think they are. They may not be strategy, they may not be knowledge, they may not be analytics or even segmentation. It’s probably a lot simpler than that.
So One of the reasons why this is tough, uh, is that we have a consumer that has realized that omnichannel behavior works in their favor. So we create traditional campaigns, we create, uh, customer journey orchestration. We have many touch points, we have many data elements. We have omnichannel behavior, omnichannel marketing, and a massive data fabric that needs weaving. And this is what I really want to talk about today, is the data, uh, ecosystem you have is preventing you from personalizing at the level you need to in order to have brand conversations, brand storytelling, proper recommendations. You have different conversations across different channels. How do we fix it? Um, imagine what level of supercharged experiences you can put in place. If you can tap into all of this data. So as retailers and consumers, which we all are here, I can tell that you are all consumers because you’re all clothed. You’ve all bought stuff, you’ve all bought, uh, homeware, wallpaper, everything. Everything that you imagine that you’ve bought, imagine the experience you have had. We are undoubtedly in a digital first environment, but the real world is where we live. And when you think of all the interactions you have as a consumer, all the experiences, all the memories of all the interactions, offline, online, in store, et cetera, across all channels and all data sources, are we really having an always on experience? Is your online experience better than your offline one, et cetera? So, um, this is where we start to talk about the gaps that we have in our knowledge of the consumer. I like to think of it as somewhere between selective amnesia and gas lighting. So if I go online and I look online and I’m looking at products and I’m having a great experience, I get recommendations.
I’ll give you a perfect example. Uh, it was very hot this year, as I’m sure you will remember. I bought an air conditioning unit. I, I spoiled myself. It was a little standup one. Um, within about three hours of actually buying that, everybody online wanted me to buy an air conditioning unit. Everyone, how many air conditioning units do you think I can have? It’s not a hobby. I want one. Even the company that I bought one from kept recommending me air conditioning units for weeks afterwards. I was tempted to have a couple, but I didn’t. So when you think about how you personalize and the relationship you have with your consumers, is it at a device level? Is it at a consumer level? Do you have a household level? It’s still between 70 and 80% of all purchase decisions made at household level are made by women. My wife does like to buy stuff and I will help her look for things, especially holiday travel, anything in the home. If I’m browsing an online, uh, uh, toaster and it’s red and I buy one, and then my wife walks in store and there is a red kettle, why would we not find a way of recommending both of those products that are similar to her at a clientele level? That’s a, a true consumer experience, and that’s what all of us really want. Without getting creepy. By the way, that’s, that’s where we go a bit too far. I know your husband bought this last week, did he? Um, so we’re gonna start with one of the most successful ways that we do engage with the consumer. We’re gonna talk about how recommendations work because when you think about personalization, but a really powerful aspect is recommendations, right?
Um, I’m assuming quite a few of you do use product recommendations on your site. People that bought this also bought that. People that browsed this also browsed that, but they work on multiple levels. And then we’ll talk about how we can supercharge those. Firstly, we talk about the consumer. The consumer. Uh, in this example, let’s call Rachel and Sarah. So Rachel will have a look at a product, um, which has already been viewed by Sarah and their behavior on their device. This is what I want you to remember. It’s device-based really looks at the behavior of are they similar people? And it.
Concept of being a similar consumer that we then use to recommend products to Rachel, not yet viewed by Sarah. That make sense? It’s, it’s incredibly simplistic in its view, and I’ll explain why It’s much more powerful than it, than it, um, appears right now. But the AI or the models are making the decisions based upon the data you give it. This goes back to data again, whatever I know about her behavior and her behavior, I’m trying to find what’s similar between them. If I don’t have all of the memories of all my interactions of Sarah and I only have some of the memories of some of the interactions of Rachel, my accuracy will be increasingly diminished. It’s fed by data. So here’s Rachel again, having a look at another product. This one hasn’t, hasn’t necessarily been viewed by Sarah. Uh, now we’re gonna take, uh, the next level of product recommendation and personalization of her experience, which is similar products. So now purely based on all the information of merchandising, like, uh, product size, color, shape, how many plates there are, et cetera, et cetera, we’re gonna match up similar products and maybe even say, I’ll recommend this product to you, Rachel, because it’s similar to what you’re looking at or similar to what you’ve purchased. Now this is incredibly powerful in its own way and years ago, this is the basic level of product recommendations and personalization that we would do. But then you have this third level on top of this, which is where more modern, uh, personalization and recommendation engines will actually take more data, more knowledge. So not just personalize the consumer and the product, but then have a look at what we know about Rachel. Maybe she is brand loyal, maybe there is a specific price point that is more appropriate for her based upon what’s been purchased before.
But the more I know, the more the AI engine can do. So if we take these two examples, uh, similar consumers and products, and then we have a look at the underlying data that fuels it, this is where I need you to fundamentally remember that, that these are our memories, right? Of all the products we have, what’s in stock? What’s, what’s, what’s the brand color and the consumers. Uh, so we have her basic facts. We have her descriptors, we have her segmentation, we have the measures and metrics that actually let us know more about her. But if we have a loyalty card, are you feeding that data in? How many times does Rachel actually visit the store? What day of the week is she a nighttime shopper? So is she always online, uh, between 11:00 PM and 1:00 AM? Is that per segment, a nighttime shopper? Does she browse in her lunch hour and then actually make a purchase on the way home? All of these memories of the interactions is actually what makes us people and we are kind of amnesia has made us forget that they are people. These are not devices. How many devices have you got? I have at least five. Excluding the air conditioning unit. I have at least five. So who am I? So this really comes down to not only are you personalizing appropriately at the consumer level, but also how are you measuring your success? If you have five different ways of speaking to me and I’m five different people and you recommend something to me to buy and I buy it once, your response rate is immediately 80% down. Cuz I’m not gonna buy it five times on every device, right? You have to have metrics, you have to have knowledge of the consumer, obviously.
Uh, more modern platforms. We then start looking at location where she was when she opened an email or checked an sms. Uh, whether for example, is it gonna be nice where she is? Is it gonna be nice and hot or wet? But ultimately it comes down to memories. The more information I can feed into this engine, the better the engine becomes. Now, what if that engine could not only engage online, but also at the point of sale, client telling applications, the iPads you have in store when you, uh, scan a loyalty card at the point of sale that happens before the credit card or the money is paid, right? That’s how you recognize the individual. So at real time, you could feed HTML recommendations and personalization to that page. Is she vegetarian, vegan? Is she ethically and morally attached to the brand? This is the consumer experience we’re trying to get to. That’s the top of the pyramid with just basic information. I’m only representing a tiny proportion of what I know about her as a person. It’s the memories and the data that are joined together. If I only have partial memories, you have consumer amnesia or worse than that, you’re actually gaslighting the whole relationship. Uh, so the concept of being able to deliver a consumer experience knowledge is king. So what are you getting wrong? How do you improve it? Maybe it’s not recommendations, maybe it’s not merchandising. Maybe it’s more than that. Maybe it’s just that you don’t all have your facts and your memories and your interactions in one place. So every channel treats your consumer as a different person. So, uh, this is a view, and I hope you can see it at the back. I tried to make it as big as I could.
Um, this is a view of what you should be thinking about as to what you have in your relationship with your consumer. Everything from, uh, in-store transactions, offline orders, returns. Uh, I have worked, I have been really fortunate to actually work with some incredible brands over the years, including, uh, next, uh, both next retail, uh, and next online. Um, Howard’s.
United Football Club. I’ve had some great experiences and each of them has their own little story to tell around actually is it a good consumer experience? And a perfect example was when Manchester United, and I’m a welshman, so I don’t like football. Anyone who hates the team, sorry about that. Uh, if you, if you remember when they launch a shirt, it’s a big event if you are a fan. So in Manchester, at their stadium, old Trafford, they had the idea of having a better consumer relationship and they put little tags in the top of all these shirts and the tag was about that big and said, write down your email address, pop it in this box, and you could potentially win a day of a personal tour at old traffic. So the marketing managers all stood there watching this amazing event as the doors opened and everybody rushed in, bought their shirt, ripped the tags off, threw them on the floor and left. Literally the floor was covered in tags. If only we can tap into what people are buying, how they’re buying it and what their experience is, we get to know the person. So do you have a device level personalization? Well, of course it’s easy. Do you have a consumer level across multiple devices? Perhaps not. What about a household level? What about a knowledge of where the consumer came from? You would not believe how easy it is to actually use the entry and exit of your website. There are UTM codes, right? We all know this. There are little codes that if you know where they come from, you can track them. What if I could tell you the influencer website, TikTok or Instagram site that a new consumer came from?
So over time, you build up a knowledge of what type of products, what type of fashion that influencer influences. So that a new entry from a unseen, anonymous, uh, website visitor from that UTM code, you immediately tap into the social network. They’re a part of. I know you from X, Y, Z. These are the kinds of things that I bought. It’s a much different experience and much more powerful. So, uh, the answer to this is, uh, better data platform, right? And we’ve all struggled with master data management and single customer views. And uh, customer data platforms been around for about four or five years. CDPs are extremely powerful to cost effectively help you bring all your data together, but they’re kind of, they haven’t really met The expectations of the market.
Has anybody tried a CDP yet? No. Yes. Part of the challenge is that the idea behind a CDP is you can throw all your data in it no matter what the data is. If it’s eCommerce, if it’s client telling, if it’s complaint systems, throw it all in. But it’s data agnostic. It’s not for retail. And therefore all the power of segmentation, modeling AI metrics, measures, recommendations, hyper personalization, it relies upon you to actually build that top of the pyramid underneath you have this wonderful data platform, but what you’ve built is a database. It assumes that you have people bright enough, skilled enough to actually build the extra point on top. And that’s the point that we want as marketeers. We don’t want capability, we want outcomes. So if you are going down a route of building a database or you’re using a CDP and you’re bringing all the memories together, make sure you build it for your vertical. You want to get that extra yard out of it, that extra mile of actually what are the metrics? Have I got a churn model? Am I making recommendations? Am I actually personalizing offers based upon arbitrary values? You come to my website, I’ll give you 10% off cause I haven’t seen you in a while. Well maybe if I didn’t give you 10% off, you would buy anyway. What if I understand more about the consumer? And this is a perfect example of next directory. So next directory. As I’m sure you know, I have got my eye on the time bus, I promise. Next directory. As you know, uh, you have to used to have these big catalogs. Anybody ever get these years ago, this big catalog, once a quarter, you’d look for it, it would weigh a ton.
It was ideal for hitting the kids with and and flicking through on a Sunday afternoon with a glass of red wine. If you didn’t purchase from that catalog, they charged you 12 pounds or thereabouts for the cost of the book as they used to call it. Um, but they had a great returns policy and the returns policy was within 28 days, you could return your product within that timeframe to either the store or next directory. So people naturally returned Unpurchased products from the store. So if you ordered product from next, by the way, this doesn’t work anymore cuz it’s been changed. Uh, you could get the book or products online, take them back to the store and never pay for the book, but never having bought a product. But next directory and next, uh, retail were different databases at that time. Totally different unconnected. One of the things that we did was Join together orders and returns.
And we found two things. One, a massive number of people would order online return in store. Ironically, you get invited to the next directory, annual sale if you order a lot. So they were inviting people that ordered a lot but didn’t actually buy anything. Secondly, we also discovered a tertiary market, which they didn’t know about. Lots of people would order from next directory, put it on eBay, and if it didn’t sell within 24 days, return it to the store. So you are powering a different consumer experience and destroying all your loyal customers. It’s not their fault, it’s just the memories were disparate. They weren’t joined together. Orders, returns were separate. So very quickly, um, we’re talking about personalized offers. Imagine what you can do if I gave you every memory of your consumers in your hand as an actionable real time insight. Every measure, no matter where they came in, how they came there, not just their device, how would you change yours? How would you change your AI driven recommendations? So for example, uh, if you buy a product, I have two daughters that, that, that are very partial to completing the look. So no matter what they’re actually looking at, these are other products that, that you might be interested in purely based on how they look, which might be length, height. I have one daughter who’s very petite and one daughter very tall. There are trousers that don’t work on my shorter daughter. She looks like a garlic trying to walk around. So understand your consumers and you can use the information from anything from recommendations at the point of sale, client selling SMS email. It all joins together and ultimately gives you this hyper personalized, supercharged experience allowing you to fundamentally change the way in which you deal with your consumers.
And it’s all down to what you remember about them. Don’t have consumer amnesia. Try and put everything in one box. I’m over E 100 with Algon, which is just over there if you want to have another conversation or chat a little bit more about this, who we’ve worked with and and how it works. Other than that, have a wonderful second day at the show and it’s been a pleasure speaking to you.
Ready for the Next in Digital-First Strategies?
Learn how to use disruptive, AI-powered technologies to create frictionless,
hyper-personalized experiences that connected consumers expect today.