E-commerce Series: Episode 3- Unlocking new waves of user growth, conversion, and retention through the power of personalized CX at scale

E-commerce Series: Episode 3- Unlocking new waves of user growth, conversion, and retention through the power of personalized CX at scale

About this Podcast

The E-commerce space has grown exponentially over the past year with evolving market trends and user behavior. Customers of our generation are not merely looking for products that satisfy their requirements but are also looking to associate with brands that consistently offer them 1:1 personalized customer experiences. 

Brands have understood this and are now looking to deliver personalization at all the touchpoints across the customer journey to ensure they are able to keep them satisfied.

To gain more insights into the E-commerce industry we have come up with this series of three episodes where we interact with global E-commerce experts and thought-leaders to gain insights into the biggest E-commerce personalization trends taking shape in 2021 and beyond. 

In the third episode, our host, Tim Moran, VP- Enterprise Engagement, Netcore and Edward Chenard, Sr. Director- Data Science and Business Intelligence at Olo unravel insights on how e-commerce brands can unlock new waves of user growth, conversion, and retention through the power of AI-led personalization. 

Edward shares insights on:

  1. How e-commerce brands can build a scalable personalization strategy by capturing data across channels and platforms
  2. The role data science plays in creating a sustainable personalization strategy
  3. How an e-commerce brand can strike the balance between using AI to deliver 1:1 product recommendations or enabling marketers to create pre-defined customer journeys 
  4. Examples of e-commerce brands that are acing personalization at scale and the strategies they implement to achieve growth
  5. Top e-commerce personalization trends to watch out for in 2021 and beyond

 

Tune in to understand how E-commerce brands are leveraging the power technologies like data science, AI, and analytics to deliver personalized exceptional customer experiences at scale.

Episode Transcripts

Tim Moran: All right. Hello, everyone. Welcome to another episode of the MarTechno beat, a specially curated podcast series powered by Netcore Solutions. Here’s where you can gain cutting edge insights from leading marketers, data scientists, product champions and Martech influencers on all thing’s user growth, engagement, retention and A.I. led personalization. Throughout this series, we’re interacting with a global ecommerce expert and thought leaders to get their insight into the biggest e-commerce personalization trends into 2021 and beyond. I’m your host, Tim Moran from Netcore Solutions, and today I’m joined by a very special guest, Edward Chenard. Edward, how are you today?

Edward: Great. Thank you.

Tim Moran: Pleasure to have you. So, for those of you that might not know, your considered is one of the top experience, science and personalization thought leaders in the world focused on driving value by creating and executing data, digital and analytics, transformation initiatives, building new business channels for large and small organizations, working along with various business lines to define the customer experience, data engineering, analytics and digital transformation strategy for the next growth opportunities, product offerings or customer experience is if you don’t have enough on your plate. You also need a broad portfolio of data science data services excuse me at Fortune five hundred and startups to include master data management, data science and repository curation. Today’s episode will unravel Deep dive insights on how pure play e-commerce and DTC e-commerce brands can unlock new waves of user growth, conversion and retention through the power of personalized customer experience at scale. Again, Edward, I covered off a lot there. It’s a pleasure to have you on the podcast. Is there anything I missed or was there anything else that those of you my audience might not necessarily know yet for?

Edward: I’m still an avid fan of playing hockey. If anybody’s into that, I live in Minnesota, so it’s the perfect place to do some outdoor skating and passing the puck around every now and then.

Tim Moran: Yeah, absolutely. And, you know, it’s interesting, an ice hockey fan, it’s something that you can enjoy in the summer, in the winter, you know, me being, of course, indoors in different rinks, but myself being from the Northeast, when I think of ice hockey, I think of snow and cold. And I just want it to be over at this point. So, please don’t take that the wrong way.

Edward: I live in Minnesota because I love the cold, so go where you like the environment. But it’s been a pretty good place for innovation since most people are indoors like half of the year because of the cold, you got plenty of time to innovate and think of new ideas.

Tim Moran: Absolutely, and you’ve had quite the experience in your professional history, you know, I’m aware of all of the corporate headquarters that are in Minnesota, but maybe you could just touch on some of the notable names that people in the audience might be aware of.

Edward:  Yeah, well, my data career really kind of started out at GE, where I was doing things like telematics and really it was my first foray into big data before that term was even on my radar. But then a few years later, I ended up at Best Buy, where I was brought in to build out personalization, and it quickly morphed into building out the big data system and then building out the data science practice and that team of mine that I built. We actually process more data than the rest of the Best Buy combined. And it was one of the pillars for the renewed blue turnaround strategy that helped us to get out of the slump that it was in at that time. Then I went over to Target and helped build out their big data science practices as well, particularly focused on the digital marketing side and trying to blend some of the in-store and online experiences. So, I got to play a little bit with the curbside pickup because I had done that over at Best Buy. And then, of course, Target wanted to pick my brain on that. So, this past year, it’s a service that a lot of people definitely have been using that they weren’t using before. Then I went over to C.H. Robinson Lathis third party logistics provider in the US and built out a product called an Atmosphere Vision, which is basically a predictive shipping engine where we help a lot of customers like Wal-Mart and Amazon figure out how to make sure their shipments actually get to their distribution centers on time, which is probably getting a lot of use right now across the middle of the U.S. And then finally, I am building out the data practice over at Ollo, which is a provider of helping many restaurants with their delivery services right now.

Tim Moran:  So,, it’s correct if I assume that you’ve had a broad experience, understandably, and you’ve had just a tremendous amount of work that you were able to do within a really diverse set of verticals from retail to restaurants and some really broad names. So, excited to get some of your insights into the topics that we have today. Specifically, you talked about Best Buy and you talked about how you handle more data within your division than within the rest of the organization combined. And I think a lot of that really, I think would lend towards scalable personalization in this case. My question is that a scalable personalization strategy for any B2C business, especially e-commerce, it’s built on a foundation of customer data, understandably capturing and consolidating that data across channels and platforms. It’s just a tremendous challenge that people face on a daily basis. In your experience, how can e-commerce brands overcome this before building out of personalization strategy?

Edward: Well, I actually think that’s sort of one of the pitfalls that a lot of companies fall into. They look at the data and they say, well, we need to capture data, big data. You know, let’s increase our volume. And there’s some validity to that. But one of the things that I went down that route myself, but one of the things I started to realize was just the degradation of the experience that we were trying to craft. So,, we went back and we really started to look at what was really important around creating a good personalization experience, and it really comes down to the relevancy of the experience, which means focusing on the data that helps you understand what is the experience someone’s trying to really have when they come in to purchase something. I like to use the example of an Xbox. Lots of people buy them for lots of different reasons. But I can guarantee you nobody buys an Xbox just to have it sit in their living room and collect dust. They’re always an experience they’re purchasing, like right now in Minnesota, it’s been like minus 15 during the day for the past two weeks. The kids are inside, you don’t want them tearing the house up, so the Xbox keeps them to contain the teenager, the 20 something year old is thinking about the games that are going to be playing against their friends. These are experiences people are buying. So, what is the experience and who is the actual individual? We have to really three-dimensional eyes our data and understand our consumers more at a personal level. And what is it that they’re trying to create with the purchases that they’re doing? That then helped us understand really what data we should collect and there was a lot of data we were just pulling in that we realized. It really isn’t relevant for what we really need to be focusing on. You know, we were collecting things like device I.D. and locations and to some extent, yeah, those things are important. But then you start to find the patterns of how people really start to search for products. So, if somebody is a deal shopper versus somebody who just wants the latest and greatest Apple product and they’re always going to wait in line to get it first, how you approach them is going to be very different. And most of the data that was being collected just didn’t help us answer those questions about how we engage them. So, once we changed that script, it really helped us to accelerate our ability to engage customers. And you’re talking about taking conversion rates from like three to five percent up until into the teens. So, it was very impactful. And, you know, when you’re talking about electronics, which is usually like a half a percent to one percent conversion without any personalization, you know, seeing conversion rates around 15 percent is pretty massive.

Tim Moran:  Yeah, I would say so, and it’s interesting, the detail that you go there about how the context of the data is so important and you need to be thinking about the shoppers and the different use cases. I mean, I think most marketers listening to this podcast do at least a certain amount of that. But really what I’m gaining from what you had just mentioned is that we need to start thinking deeper about the data. We need to understand how that data informs our view of the customer and then leverage that to better curate the experience for the customer, not only in an in-store kind of a situation, but to understand how they’re going to use the product. Right. Like you mentioned, if there somebody in Minnesota looking at negative 15-degree days, what’s that product going to be used for? I think a lot of the press about Xbox and PlayStation these days is about the reseller market and how people just can’t get a hold of one. And it’s interesting to me how retailers have used their data to overcome that challenge by bundling products. Right. And putting a bundle together with games and controllers and things that it doesn’t make a lot of sense for the resale, the resale person to go and, you know, basically go and try to scout that. But it makes a lot of sense for the shopper because they’re things that consumers would have purchased anyway. So, I mean, in that regard, there’s so much data that not only electronics retailers, but fashion or home appliances, any type of retailer, basically, or many different types of digital businesses. The depth of data is incredible. And I’m just curious about what fees do you have about the rules that data science and what it plays in creating sustainable personalization strategy?

Edward: Well, for me, data science is there, there’s a kind of doubt, the standard or now we can even probably call it the traditional view of what a data scientist does, which is focus on stats coding. That I actually found to be quite limiting and a good example. One of the companies I went to, they were trying to do personalization with a lot of statistician’s data scientists, and it just wasn’t working. They really weren’t connecting with the customer. And one of the things I asked was, so how do you determine your algorithm is right for the market? But we just listen to the data. But do you understand the customer that’s creating the data? And they really didn’t understand where I was going with that, so right away I brought in the persona’s that had been developed by design, said which one of these personas is going to use that particular algorithm and at what point in the customer journey? They couldn’t answer that, so I started bringing in people that have backgrounds in anthropology, sociology, psychology, we actually started scrubbing the biases out of the algorithms and focusing on the biases that were much more what you would see with the personas that got the conversion rates up fairly quickly. And since then, I’ve really had a mixture of data science skill sets where there is more of that behavioral science piece, but also that traditional statistical approach as well. And that’s something I still find is fairly unique in the marketplace. You just don’t see a lot of behavioral scientists on data scientist teams, which I think is a real shame. They definitely should be. And even using linguists just to regionalize your message. So, how we speak and Minnesota is not how people speak in New York City, languages a little less direct out here than it is in New York. Having a linguist help to craft the regional message. You know, you can still have a template email by region, but having it really sound like someone who is like your neighbor is more impactful than just getting a canned email. That sounds like a Minnesotan talking to a New Yorker. And I can tell you the impact of that when I initially did an experiment around that. We ended up getting a 400 percent increase in open and conversion rates on emails, and we have to run the experiment twice because I was like, did we screw up our math somewhere along the way? This can’t be possible. How do we get this good? And we kept running it. Sure enough, we kept hidden in that range. So, it really demonstrated to me the impact of bringing in behavioral science and social sciences into the mix.

Tim Moran: That makes a tremendous amount of sense and, you know, it’s interesting from a large grand standpoint, the resources that are available to be able to bring in data scientists and bring in behavioral data scientists, many small organizations don’t necessarily, I think, have the budget or the luxury to be able to invest in those different types of specialists. You know, we see that a lot at Netcore ourselves. We often work with companies to help drive engagement rates and email rates for open. And it often comes with a balance of, you know, a mix of software and services or artificial intelligence and, you know, marketers using predefined journeys on a customer engagement platform. Know, if you think of from one end of the spectrum, using artificial intelligence to power one to one product recommendations and journeys, and then the other end of the spectrum is a marketer that’s using predefined journeys that they’ve thought through. And maybe they didn’t have a behavioral data scientist, but they’ve tried to put themselves in the shoes of the shoppers, as good as possible as well as possible. How do you think that e-commerce brands can strike the balance between those two very different approaches? Maybe that, again, they don’t have the resources to have all of the data scientists available?

Edward: Well, what I find so large companies, you often will have very good technical talent available to you at smaller organizations that there may not be the case, but where smaller organizations I find often excel is you do have a number of people that have far better expert intuition about their market than you see on a larger organization. So, you may have somebody who’s been with the company since they left school and been there for 15 years. They probably know their market really well at that point in time. Whereas at the larger companies, you know, particularly on the data science side, like the average person’s in and out like two or three years. So, they’re sticking around long enough to develop that expert intuition. If you have someone on your team that does have that level of expert intuition, I would actually say that that’s probably stronger than the actual algorithms being built by the larger companies. And it’s just been my experience that. As long as they’re staying relevant and current with the market, you can really do a lot because what really is, is an algorithm. It’s really a collection of assumptions and biases that somebody baked into code. And. Trying to replicate that expert intuition on a mass scale. That’s the one drawback if you do have those couple of people at a small company, you really can’t scale them up that easily. Now, if you can replicate their knowledge into an algorithm, that would be great for helping you to scale. And there are some methods to help you do that. It’s just not something that most small organizations are even going to become aware of because they just won’t have the technical talent usually available to them or even the understanding that there’s things like swarm that they could actually try to harness to gather that expert intuition into an algorithm to help them.

Tim Moran:  So, you’re talking about biases and A.I. and I immediately think of I think it was the example of was it was Microsoft that put a chatbot up, really open to the Internet and it went off the rails very quickly. And I think and again, please correct me if I’m wrong. Sure. I’ll get an email about this if I miss naming the company. So, again, I’ll have to double check that. But you know, the thought there is use people and trust people who know the space, trust the consumer or trust the person that has experience with the organization that understands your consumers, your shoppers, and use them to not only create and craft and curate those customer journeys, maybe in a predefined fashion, but in the cases where maybe there is some level of artificial intelligence being leveraged within the organization, maybe almost use them. Is a pseudo data scientist, a pseudo behavioral scientist to spot checks for the A.I. to prove the results and to understand if any type of intervention or tweaks need to happen? Would that be a correct statement?

Edward: Yeah, you know, domain knowledge is to me very crucial. And for my team’s data scientists and engineers, I always have them target Best Buy. Go to the stores, spend time working in the stores, understand what it’s like, really observe how customers are using their devices in the store to get a better understanding. Because if you’re just sitting in your own little cubicle or office, you just can’t really rely on the data. Because, again, I look at data and I think that most companies misinterpreted it as being something concrete, very logical. But I look at it as very interpretive. So, unless we’re looking at ones and zeros, I’ve ever really seen data. No, you’re really interpreting it. Most of the time. It’s either in a spreadsheet, it’s either on a Web page or PowerPoint. That’s all data just being interpreted in a specific way. So, the interpretation of it really impacts what people walk away with in terms of the knowledge they gain from it.

Tim Moran: Yeah, that’s so true, and it’s the interpretation, again, that can lead to bias, but if you have those trusted people, you’re more likely, I think, to potentially get a better outcome in that case if you go to the other.

Edward: Whenever I hear the phrase like I just got into that, I always think that’s the total cop out because data doesn’t speak. It’s just interpretive. How well did you interpret it? That’s really what I’m usually drilling down into when I look at any type of algorithm. What were the biases and assumptions you baked in there? And, you know, some of them that the teams, though, they’ll design are really based off of their own shopping patterns. And it’s like, are you the primary market? So, a good example at Target, you know, 80 something percent of the target regular shopping base is female. Some of the teams that developed the first albums and data science and personalization were all male and they really didn’t go shopping and they didn’t really shop like the average shopper target. And so, there were a lot of biases and assumptions baked into the algorithms that were not accurate. And because it had to be changed.

Tim Moran: It puts a lot of context on it, too. And how much we trust interpretations of data, who’s providing those interpretations? What’s their experience with the relevant consumer set? That’s really interesting stuff to be able to consume, I think, in the future in terms of future learning or things to observe within a marketing program. Now, if you take yourself and put yourself into the shoes of, let’s say, the mid-market business and they’re looking to grow and expand and become more sophisticated, maybe you can provide some examples, maybe like two or three examples of, you know, e-commerce brands besides like the pioneers of Amazon and the others, but maybe some great examples of brands that are facing personalization at scale and some thoughts on what they’re getting, right.

Edward:  You know, I really liked renting the runway when I first was introduced to their brand, just the way they allowed people to take photos of how the average person looked wearing an outfit. And I thought it was great because instead of just getting some model which has like zero percent body fat and, you know, it looks great. And now you can see somebody that actually looks more like you and how they look wearing that clothes, you can be like, oh, yeah, this is probably how I would look on me too, since we have similar body shape. I thought that was a really great feature. And then the names escaping me right now, but there’s a company down in Brazil and they’re a clothing company kind of mid-market size, Brazil’s an interesting one because a lot of the personalization that we’re allowed to do in the US is illegal down in Brazil. So, a lot of the tracking that we can do, they can’t their method of getting around that was to allow people to plug in their own measurements and have the clothes selected for them. So, basically curating the catalog just based off of their own measurements and other preferences they were putting in there. And I thought it was a very good, creative way to get around some restrictions that were placed on them that here in the U.S. we just don’t have to deal with. So, I thought it was pretty good, actually. Overall, I thought there were a lot of companies down in Brazil and other places like Chile and Argentina that were quite creative in their approaches versus what we do here in the U.S. So, if you want to look outside of the U.S., I would actually look at Latin America as a place for some innovation happening in the e-commerce digital space. And that’s interesting. You know, I would just say. You know, some of the more midsize companies like Maury says that that’s a Minnesota brand up in Duluth. I think they’re also doing some pretty good, just moderate changes along the way of just trying to increase personalization, but not disrupting the experience too much, because that’s sometimes what I see as you can implement these new programs, that it’s almost like changing your brand too much, too fast. And you can lose some customers along the way, whereas they were taking much more of an incremental approach, really changing the brand slowly over time. And I do think that that’s a good approach. Even though I like to see rapid change, I know I’m probably in the minority when it comes to the whole of the population. So, they’re being more conscious about what that level of change and adaptation for the consumer actually is. And I think that’s a good approach.

Tim Moran: Yeah, and I think that’s probably a lot more attainable, too, for the audience that we’re talking about, for the middle tier business audience, you know, the idea of incremental changes where you can change, make a change and then observe the effect that change has on your business. Really interesting point that you brought up, too, especially about looking at the Latin American market and about the restrictions they deal with. You know, at Netcore, we have a global base of customers. And, you know, we’re often speaking with e-commerce, retail and e-commerce specific companies and then brick and mortar and e-commerce retailers as well as others. And there are a lot of data challenges, so much so that we’ve developed our products with the thought of making sure that, you know, they can stay well within the regulations of the various localities that they’re in. Like our personalization products, for example, we limit the use of PII. But overall, you know, in the sense that, like you mentioned, if you can look at the innovation that’s done within areas where there are limitations in the creative ways, people are overcoming that, that’s often I think now that you mention it, a really excellent place to find some nuggets that you can pattern yourself or model yourself after. So, I think we’re running down on time. One last question for you, and I like to ask everybody this. So, if you look into your crystal ball, what are maybe one or two top e-commerce personalization trends that you think people should really watch out for in the coming year?

Edward:  I think the more people are becoming more cognizant of the value of their own data. And there have been some discussions around like personal clouds where people can manage their own data a bit better. I think that’s probably a more of a long-term thing that’s going to happen. But as the younger generation, they seem to understand the value of the data more. I can see that happening more and more over the next decade. The other thing is just really moving away from that templated approach, personalization. Even if you go to the large companies, it’s still pretty much just recommenders and slapping someone’s name on a templated email. And that’s just table stakes. If you really want to engage people, it’s changing the dialog instead of just we’re going to. I remember once I was interviewing with a large brand, I won’t say who, but they said we define the experience for our customers. And I was just thinking, wow, how like last century that thinking is. It’s really the other way around creating an environment for the consumers to tell you, hey, here’s what I like to create, help me make that happen. It’s actually a much more rewarding approach than trying to figure out what the trends are. You’re really creating a dialog with the customers. And at the end of the day, if somebody has a conversation going on with you, they’re going to stay with you. They’re not going to look at you as just a commodity. When I go onto Amazon, it’s really a commodity. And a lot of ways it’s like, well, what’s the lowest price I can get here? Maybe I can go to Wal-Mart, get it lower, but if I get a relationship going, it feels like a real relationship. I’ve actually got back and forth going forward, going on like, let’s say I wanted to. I’m looking for a new mountain bike once the snowball melts. Love to have a company that actually engages with me, understands the experience I’m trying to create here and really builds some type of connection. And actually, when I was doing consulting, I worked with a company out of Chicago. They do bike accessories for road and mountain bikes and that’s exactly what we did, created this connection with the brand and the experience that the average person wanted to have. So, instead of focusing on the hardcore user, it was really much more the weekend writer who wants to attain that, that look and image that they’re kind of like the constant biker. That’s really what helped engage with them, they were able to get there, the number of new customers above the forecasted rate that they were looking for with their new products and their new app that they had launched. I do think that’s really where it has to go, figure out how do I create a good conversation? Because at the end of the day, if you don’t have that, you’re still just guessing. And that’s really what I understood when I was at Best Buy and Target. If you follow the traditional personalization approaches, you’re really just throwing stuff out there saying, hey, does this interest, you know, OK, what about this? All right. I’ll recommend this other product here. Just get a conversation. Just ask, hey, why, why? Why are you coming to engage with us today? What is it we can help you with? It’s going to give you so much better information and get you to that sale so much faster. I mean, when I was a Target Best Buy, we saw research that said it takes people will look at ten different sources before making a purchase. If you got a relationship, they’re probably just going to deal with you or maybe one or two others at the most. So, your odds of getting that sale are going to be much higher. And I bet you the conversion rate and the cost of acquisition and retainment, it’s going to go down across the board.

Tim Moran: Yeah, that’s excellent. And Edward, just I mean, to cover everything, I think the way that you summarized that really pulls together just about everything that you’ve spoken about during our time together here. You know, there’s value in your own data as an organization. The goal is to create an engaging experience, not just understand data or technology or software for technology’s sake. Really, the goal should be to observe consumer behavior, to better understand that data and to understand if there aren’t biases in the data in the way in which you are representing that to consumer behavior. And then last but not least, it comes down to the people, whether you’re a large organization who can afford behavioral data scientists or your other businesses that maybe don’t have that luxury. There’s a real value in trusting people who have experience in this space. And it’s those people who know the space and understand who the consumer is and what their value system is. That can better help you to either work with those predefined journeys to observe if there’s behavioral biases within data or to understand it as maybe not performing in the way that it should. So, in summary, Edward, thank you so much for joining us and for making the time today. I definitely appreciate you speaking with us.

Edward: Yeah, no problem, and, you know, I guess there’s just one final thought, I would just say that I often get called the data scientist and I was correct. Everybody that does I never call myself one. I really think in particular for a marker. I actually came out of marketing my degrees with marketing. My first job was in marketing. I like to say I’m more of an experienced strategist because at the end of the day, I think that’s the real focus. It’s not the data. The data helps you understand the experience and how to engage, but it’s really about crafting the right experience for the right audience.

Tim Moran: That’s excellent. You’ve given us a lot to think about, and I’m sure that our global audience will benefit greatly from everything that you shared with us today. You know, again, Edward, if they want to follow up on more of your thoughts, where’s the best place for people to find you maybe to contact or through social channels?

Edward: Really, the only place I generally post on a regular basis on LinkedIn, so I just just look me up on there, I mean, I usually always accept a connection request.

Tim Moran: That’s great, and again, for everybody, it’s Edward Chenard. Look him up on LinkedIn, he’s the one with just the tremendous amount of experience. You probably can’t miss him. And again, a big thank you to our listeners for tuning in for more content on all things’s personalization, audience science and marketing technology. Do subscribe or follow the Martechno Beat on netcorecloud.com or write to us that [email protected] with any suggestions you might have. Again, my name is Tim Moran and thanks for listening today.

Unlock unmatched customer experiences,
get started now
Let us show you what's possible with Netcore.