The Power of E-Commerce Personalization with Dr. Kyle Allison

In the world of e-commerce, forming genuine connections with customers can feel like a challenge. By harnessing the power of personalization to delve into consumer behavior and preferences, businesses can uncover the triggers and influences that shape their buying habits. Doctor of Digital Strategy, Kyle Allison, is here to help us brainstorm ways we can create human connections with our own customers. 

Key Takeaways: 

• E-commerce personalization extends beyond product recommendations, focusing on customer attitudes, emotions, and decision-making. 
• Utilizing tools like chatbots and surveys to gather first-party data enhances personalization efforts. 
• Effective personalization strategies require regular assessment and adaptation, evolving with your customer base and business growth. 

Rich: My guest today brings over 20 years of experience in e-commerce, business strategy, digital analytics, and marketing. He’s held leadership roles at companies like Best Buy, Dick’s Sporting Goods, Dickies, and the Army and Air Force Exchange Service. 

He’s known for his innovative strategies in digital marketing and e-commerce, which are based on data driven insights and apply to various business sectors. In the academic world, he’s a professor and mentor, teaching subjects like digital marketing and business strategies at respective universities. He has also contributed to academic literature and has a strong educational background with degrees in business and communication. 

Today we’re going to be looking at how you can create humanistic connections in e- commerce through personalization with Dr. Kyle Allison, aka the Doctor of Digital Strategy. Kyle, welcome to the show. 

Kyle: Thank you, Rich. I appreciate that. And I’m so glad to be here today.  

Rich: Okay. So personalization means different things to different people. How do you define it in this context?  

Kyle: In this context of today’s environment, personalization is more than just the offer and the value of what you’re offering. It’s the connection that you know the customer more intimately, more about their ethos, their emotions, their attitudes. I look at it as a consumer behavior kind of class. We teach in that class – attitudes, behaviors, the decision-making process of what goes through their minds and their beliefs.  

So I think in today’s environment, personalization is more than I just know what you like. I know what you prefer in your products or services. I’m learning more about your behavior and your habits. What makes you want to buy, what triggers you and influences you. That’s personalization truly in the environment today.  

Rich: So whenever I think of e-commerce, Amazon’s always one of the first companies I think of. And I think that’s one that we’re all familiar with, even if we can’t quite replicate their business model and their success on our own websites. But are there examples from Amazon that you can pull from so we wrap our heads around what level of personalization might be achievable?  

Kyle: Yeah. So Amazon, obviously, we log into Amazon and we have history with Amazon. They’re going to show you on their homepage, front and center, products that you’ve looked at or bought before, and similar brands and products like that.  

What they’re doing fairly well, I’m sure, is modeling beyond that time and place. When did you look at this seasonality kind of things, they’re not going to have you look at something for outdoor apparel in the middle of, let’s say, winter, that kind of thing. They’re going to do it at the right time, right season. And then we look at your history from back last winter when you bought winter accessories and bring it to you at the right time to buy that. So they’re good at that.  

I think when you look at Amazon or a major website who really understands personalization today, they’re looking at what did you do before the purchase? What was your path to purchase? Really leveraging Google analytics or Adobe web analytics and really understanding the path and what you took along the way.  

But there’s an added value though, to something that we’re going to talk about in a minute, on building your own data set beyond just the clicks and the views and the purchases. That’s something you’re going to get, I call it ‘descriptively’. You’re already getting that with the actions your customers online take place. But we have to do something more than that to understand more of these attributes of our customer behavior. Which is what we’re going to talk about here in just a minute. 

Rich: So let’s talk about attributes. So what do you mean when you’re talking about this? What is beyond my purchase history when I go to an e-commerce store? And if I’m running an e-commerce store, what things can I start to pay attention to so I can deliver a better experience for the people who come to shop with me? 

Kyle: Absolutely. So attributes are really your characteristics of how you define your customer profiles. We already know enough about them as far as what they bought, what they viewed, what they clicked. And we know that views and clicks do show interest and intent. That’s the classic model of personalization and just tracking customer behavior online. What I’m talking about, though, is tapping into the belief systems, the attitudes. Like truly do they value us as a brand or company or not? 

And truly more on their interest level as well. We can give surveys, right? Sure, that’s classic. Give a survey. Ask questions. But that’s not really the best way of doing it in today’s environment. If you develop tests of personalization, you develop a forum for where you’re actually communicating with them and learning directly about those attributes. Which are characteristics of your customer, but you need to define it for your customer. However you want to define it, whether it’s a passive shopper we call a loyal customer.  

What does loyalty really even mean? Are they loyal because they love the brand, they love the product? We’re just good to them? And you can start taking a lower-level step in attributes and really understanding how to code them, if you will. When I say ‘coding’, not programming. Label them in your data the way you want to define them.  

I can give an example. An online footwear retailer I know of actually use their chatbots as more than a service like handling customer service issues. They went on there and actually had a conversation midway through the shopping experience for their customers – I call it ‘midpoint funnel’. As they started tracking the behavior of a user on their website and what shoes they were looking, maybe brand, style, they’d have a chat bot pop up called “footwear fashion friend” that said, “Hi, I see you’re searching dance shoes. Can you tell me more about what your event looks like?” And they actually started in a dialogue with some chat bot AI asking them what type of dance are you going to? Oh, it is an outdoor event. Here’s some shoes we recommend for it being outdoors. But through that dialogue, if you will, we’re gathering data. Now we’re curating our own data and building our own data sets around the behaviors of our users based on what they’re telling us via chatbot.  

That’s one medium to do that. And there’s other ways to do it. This is an example, but through that dialogue we may learn about, hey, they actually prefer a brand that has sustainable shoe fabric. So maybe the conversation is, what else is important to you about these pair of shoes? The customer may say, “I want environmentally friendly.” They recommend not just the product that serves that, but now we can label that customer under our environmentally friendly, conscious customer, sustainable customer.  

This is what I’m saying. You can build these experiences and then learn how to absorb more information from your customers, and then build that data to your advantage to take something called hyper personalization to take it to the next level. So that’s an example of things you can do.  

Rich: So as you’re describing this, I could see when companies are first starting out, we’ll just stick with the shoe company. So a shoe company might say, okay, do we want people who are more interested in outdoor gear, or just dress gear, or business gear? And that may be some of the codes or the tags that we start to apply to our own database of information. Our first-person data, if you will.  

But then we might even get more specialized over time, or maybe if we’re focused on a specific niche. Then we start doing even more of these tags and we start to get information, if I’m understanding this correctly, not just about that individual who we can better serve, but we’re also getting a better understanding of all of our clients and may be able to provide better recommendations to them as well. Is that correct? 

Kyle: Absolutely. So we can tag, like you said, the straightforward actions they take on a website, the events, if you will, as Google classifies it, of a click or view. But if we get more with the dialogue in the chatbot or more of a form of a webinar, we’re engaging their customers. We get more of a yes. It’s a sample. We get better at understanding their lifestyle, the needs of their products beyond just they wanted because of a good price. Maybe the durability and the look of a pair of shoes.  

There’s more that there’s behind a customer’s intent, but not all customers. We’re going to understand. Everything about them, but we can get better and learning more about them through these curated digital experiences. And a lot of it is, let’s just say it’s a first-time website. i’s just learning how to do personalization first. Don’t get me wrong, you have to set up the tool, learn the basic mechanics of personalization, which is tracking views, clicks, purchase history, and making recommendations from that. But once we’re really good at that, the next step up is let’s do some testing on some curated of our own customized experiences, and build a forum to learn more about these extra layers of interest for our customers. 

And really what you’re doing is you’re adding more value to your data to return better experiences moving forward. And absolutely, once you have the right sample size of customers you’ve learned from, you can turn that into full blown strategies at that point. Personalization is a place to go one to one, but you can also learn from that how to make things better all in over time. You can take further marketing ads on the side or what to promote better to, and not just that time and place for that customer, you can build long term campaign strategies based off the learning of that personalization you’re doing.  

Rich: Now we’ve talked about Amazon a few times, and obviously I doubt that the lead programmers at Amazon are tuned into this podcast, but maybe they are. “Hello.” But for most of our listeners, I think that they’re running their e-commerce shops on platforms we’re all familiar with, like Shopify or BigCommerce, WooCommerce, what have you. Are those tools robust enough to do this, or are there add-ons or plugins that you have used that you recommend that would put this type of technology and personalization in the hands of the people listening to this podcast? 

Kyle: So the fundamental personalization, some of these platforms that you’ve mentioned do some basic recommendations. And it’s probably going to be, like I talked about the fundamentals of if the customer saw this, they also bought that, or it’s a basic rule setting.  

So the thing is, it all goes down to the very narrow, the types of rules you’re allowed to use as the e-commerce, as the store owner, whatever you want to call it, the person who’s going to be behind the scenes, what do they have access to the business user? And it’s all about the number of available rules you can set for your personalization.  

Some of these platforms will have either maybe an add on service for a little bit more of an investment, but it can be built software out there that are specializing in this. Like Monetate is a big company that does this, they’re topping the game. Optimizely is another one, AB Tasty. So there’s a few there I would say can do more. They have more rules you can do. They actually have other types of data built in, like weather built in as an example, and seasonality of your customers or where they are based off, whether they have other things like one that actually tracks the stock market. How does that impact customer behavior around a certain industry, customers react to certain ways?  

So there are some other variables they can bring in to help you, hyper personalize this at the next level. So I would say they’ll get familiar with what’s already available to you. First, learn the fundamentals of personalization there before you go invest into the, let’s say that in the upper echelon of personalization tools, unless you’re ready to go. 

Once you have some history, there is some out of the box, I call it. But there are also tools out there like Monetate, Optimizely, who specialize in this at the next level.  

Rich: And these tools that you mentioned, are they e-commerce platforms unto themselves, or are they more like they integrate with the platforms that we may have already set up our stores on? 

Kyle: They’re going to integrate, for the most part. Monetate is owned by, I should know this, they have their own commerce sister company, if you will, under their platform, but they add on to any e commerce platform out there.  

So the commerce part is what we’re talking about, like a Shopify and WooCommerce and all that. These are the personalization tools. They will say, we have our own commerce platform. You don’t need to have that though to do the personalization, they integrate with it. But you run all your personalization on their tool.  

What’s good about these tools though, too, is they have a staging area where you can view what you want the experience to look like before you hit the trigger. And a lot of them have inside of them, the analytics themselves, right? So you can plug it into Google analytics if you want, or whatever platform of analytics, but they have their own set of analytics.  

And what it comes down to is you can actually build customized rules. So the value of these add on tools is, like I said before, you may be limited on the ones that are just built in or the basic personalization. These other ones allow you to curate very specific rules. If you wanted to add in multiple layers of rules like, “This customer who is price conscious only when they see a price between, $29.99 and $39.99, promote the product.” And it’s very specific.  

Like for a price as an example, anytime our price is above $39.99, do not show that product. But if it’s on promotion, it triggers an automated personalization, then promote it, right? We only know those customers shop between these hours of the day, and only when these brands are showcased. Get more specific around time of day. Brands and the rules are way more granular, which is good if you have a very strong mix of products and a lot of diverse customers, obviously. If you have a very straightforward brand with a small set of products and a very small segment of customers. What I mean by small, not scalability, I’m talking characteristics, your personalization is more of a good feel, right?  

So personalization isn’t just promoting products. It’s how do I connect with you on a brand level? We call it brand love. How do I make sure you feel like, “Hey, Kyle, you’re on the website. How are you today?” There is that element, which is part of this. But from a true sales perspective, it’s getting very specific on ensuring the right attitudes are understood by your customers. Even like things like what do they have interest in as far as maybe politics, as an example. You could try to figure out a way to say, products serve what political views of a customer if you have that savviness. So it gets down to that level of detail for the most part.  

Rich: It sounds pretty customizable. You use the word ‘granular’. So I’m just trying to envision in my head as we’re gathering this data and I’m using these tools, it sounds like I’m assigning tags like #ecoconscious, or #sociallyconscious, or #budgetconscious.  

I assume there’s also some preset ones that you can do, but am I manually applying these to different shoppers? Or is this more of an AI play where it sees the behavior and it makes some good guesses and then I can correct it if I don’t think it’s on target?  

Kyle: It’s going to be most of the time you need to define it yourself first. What are the parameters? And then when the parameters are met, then those customers who follow that parameter, the AI will absorb that and then provide the possibly recommended experience.  

And it’s not just products. It’s the banner. It’s the ad, whatever. But you need to define up front what those characteristics are. AI can definitely bring you back the results of previous experiences, for sure, of what you’ve done before, and you can assess it that way. But when I say ‘granular’, the only reason granular is you’re not really doing a lot of manual work. You are building though granular, more defined, if you will. It’s probably better for a defined set of characteristics about your customers or the experience that you want to understand better. So that’s the more narrow part. 

But once you have that done and it’s tagged and all the absorption around that, any online user who fits that set of characteristics – we talked about attributes – they’ll be served that experience. Which is good because you want that. You want to know even newer customers coming in who fit that criteria so they can be part of that segment for the most part. 

Rich: You mentioned banners and ads in passing. And so it brings up a good question. Obviously, these things will impact the experience once somebody is on the website. But is there a way to optimize our ad campaigns or email marketing campaigns using these same tools, or does it start and stop at the website? 

Kyle: I’m actually really glad you asked that. One of the things when I teach and when I consult on e-commerce, is your e-commerce website – in general, this is about advertising in general – personalization tools or not, whether you do it on your own analytics on or without AI. Your website has so much rich information about what your customers think, feel, say, even if they use onsite search and all these other things that the functionality website can do.  

You need to absorb that data and that information really, and insight and bring it back to the outbound marketing, like you just talked about email, social, whatnot, Google ads, because that’s already what the people on your destination are saying. I feel like we focus so much on the top of the funnel, what’s working, not working out there, between the different channels, which is what you need to do. You need to manage your outbound channels, but we don’t take the insights or what’s happening on the website to say if it works here on the website and we have all this information about what they prefer, what they do, what’s triggering them to want to buy or not buy. 

Obviously, we want to use what’s triggering them to buy, bring that back out to the fold, send emails out there to that same customer group. If they’re logged in, of course, you know they are, on personalizing your emails that way. Or take a shot if you have enough of a sample size, to a new audience thinking they fit the same set of characteristics and segmentation, so yes, absolutely bring that back in. So yes, the long story is yes, you can.  

I just want to reiterate though, I feel like in general even if you’re not using AI and personalization, you can take your descriptive track record of what’s happening on your website with all this user behavior. Even with the general Google Analytics view, you get on your default view and say, “Well, I have something I know they click here. They view these pages of that content that they seem to like that they’re converting from.” Bring that out because more people will see that.  

And obviously we know that with the quality score of Google ads or just search in general, the click has to make sense to the link. That’s fundamental. We understand that as digital marketers, but for those who don’t understand, that has to be done. Your landing page for e- commerce has to match whatever you put out there for Google to say, yeah this makes sense.  

So yes, long story short, yes. It’s my passion subject, actually. How do you connect the inbound and outbound strategically? And that’s really learning from the inbound more than just the outbound.  

Rich: All right. You mentioned that you can just start with some of the basic things that are on the platforms that listeners may already be familiar with. How do you know when you’re ready to add an Optimizely or Monetate to the website? 

Is there a trigger that you might recommend when you get to a certain point that this is when you should be investigating or implementing some of these more higher end personalization tactics?  

Kyle: It really goes down to your level of operational maturity. Are you really using personalization as a regular tool, strategic planning tool? Are you looking at it every now and then? If you’re just ad hoc, then I would say you’re not ready yet. You need to make a commitment, it’s really about that. Are you integrating this as part of your ongoing business review process, development, strategic planning, whatever you want to call it. It’s got to be part of your, day to day, week to week mentality.  

Now personalization, of course, up front takes time for the results to really showcase itself. And personalization is not a campaign like, hey, one day campaign, you’re going to see all this data and understand it. Because personalization is based off a score, which is a metric that you’re evaluating, experiences take time to marinate. So you have to be patient with personalization.  

First off, understand it may take a week, two weeks sometimes, for a test to truly get a confidence level and say, yeah, we believe in this test or not. It’s working. It’s not working. But otherwise, yeah, the investments on you if you’re really going to do it. So that’s one number one.  

I think that just your maturity as an organization to say, I will not just invest into it, I’m going to actually use it. It’s also if you’re getting some good results off your foundational personalization. What I mean is, have you seen something convert better? Have you seen some growth in some metric from the fundamental personalization you’re doing? If not, I would say keep trying to figure out the fundamentals first and see where you can make some wins. Because if you’re not winning in the fundamentals. It’s going to be harder for you to understand how to win in the advanced stuff. ‘Stuff’ is such a general word, but that to me is what’s important.  

And then from there, you have to have a decent customer base of returning traffic. If you’re not growing your traffic organically, I would wait. Because organic traffic, even a little bit of organic traffic, is what you need for these personalization tools to work. Because what happens if you’re not growing your organic traffic or growing your traffic period and new traffic, your retention of personalization becomes saturated. It’s the same mix of customers with the same behavior. So hopefully you have somewhat of a growth plan on top of using personalization on the website. Now you can use the personalization, like I said a minute ago, with you to bring that back out to grow new traffic. But you need somewhat of a foundation of growth. And that really depends on the numbers and all that.  

But if you’re not growing, you’re not going to learn new personalization because you’re going to have the same regurgitation of personalization, the same customers over and over. And at that point, it’s how often are customers going to need to buy from you anyways? So it’s a mix of both traffic levels and how you’re going to implement it and use it.  

Rich: It seems like this could take quite a bit of our time. I could definitely see going down a rabbit hole with some of this. What do you think is a recommended amount of time that somebody should expect to be working on these tools, like a range each week or each month? 

Kyle: Yeah, so if you are a small team privately on your cell, you’re one person or small team of less than 10. You need to have someone devoted to it first off, right? Someone needs to be that, I call it either personalization or a B2B specialist. Someone needs to be the owner of the tool like anything else, right? So define that person. Otherwise, no one’s going to know who’s doing what and it gets mixed up in all that organizational stuff.  

I would say to start, you really only need a few hours a week. I think to learn the tool, it may take some training, couple of weeks or so. I think some of these companies should provide that or offer it to you. So work with them first off on their training plan. It may take a couple of weeks or so to be trained on it.  

But to start some strategies, just set up a handful. Get a few out there. Three, four, five of them, and just let it do its thing for a couple weeks or so. It may not take that long for it to actually say, hey, you’ve got a winning score. If you have enough traffic, it’s all about traffic. Again, that’s a big deal. You need the population to learn from. But anyways, a few hours a week.  

And then if you get really good at it, some websites run dozens, if not maybe hundreds, of tests on the personalization strategies at one time. You’ll need a full time resource at that point to do this day in and day out. Maybe they have other responsibilities, but you want to monitor your results weekly. You’re not going to do a lot of day to day for changes of your tests or personalization. So you’re not really saying every day you need to spend hours on it, but you need to assess it at least once a week and then make changes as you see fit. 

Sometimes it could be in that week when you look at the results of one personalization versus another. I maybe pivot it or I don’t. You let keep it going. So upfront, a few hours a week, three to four, maybe. And then over time, it could be someone’s part time job because they’re doing more tests, more evaluation, more decision makers that need to become involved. 

So the more you do, the longer it’s going to take. So upfront, a few hours. After that, it really is a resource. Someone’s going to have to do it part time or full time.  

Rich: All right. This has been really enlightening. Kyle, if people want to learn more about Dr. Kyle Allison, see what you’re up to, follow you, connect with you, where can we send them? 

Kyle: Sure. Doctorofdigitalstrategy.com. Everything’s there.  

Rich: All right. And we’ll have links to that in the show notes. Kyle, thank you so much for your time today. Really appreciate it. 

Kyle: Yes. Loved it. Thank you so much. 

Show Notes:  

Dr. Kyle Allison is known as the “Doctor of Digital Strategy”, thanks to his passion around all areas of digital marketing. He educates others on how to drive value through building effective digital marketing strategies with a human approach.  Check out his blog for helpful articles on SEO, Customer Value Proposition, AI, and more.  

Rich Brooks is the President of flyte new media, a web design & digital marketing agency in Portland, Maine, and founder of the Agents of Change. He’s passionate about helping small businesses grow online and has put his 25+ years of experience into the book, The Lead Machine: The Small Business Guide to Digital Marketing.