Close the Experience Gap with Proven Personalization Capabilities

At this point, the term personalization means everything and nothing. We all know personalized customer experiences are all that matter, but many brands still struggle to deliver these at scale. Watch Ryan Greene and James Meyers as they unpack what it takes to actually create the type of experiences that drive customer growth and loyalty.

Ryan Greene: 00:03
Okay, so let's get into the topic. And I guess I'll start by saying the topic today is personalization. So here's to that. And before we go too deep, why don't we just explain, go a little bit into why we're talking about this. What is it, why is it important? And then once we understand that, we can go into the different aspects and dimensions of this that might be interesting or confusing to a lot of people who are listening today. So, why don't you start with - what is it it? And "it", you can say or not say, but what is it?

James Myers: 00:44
Yeah, from a high-level perspective, if you think about personalization as helping consumers...

James Myers: 00:53
It's about helping consumers achieve their goal, right? And you can, you can personalize websites, marketing, commerce experiences...the list goes on and on. But the reason why it's so important is because at first, if you start with your customers, there was a recent statistic that came out said 72% of consumers will only engage with marketing if the content is relevant to their interests. So that's 72% will only engage. So, now we cover the customer and why it's important, but let's think about from a monetary value, right? I think from my experience as a practitioner, and you can tell us yours as well, the common lift that I've heard is somewhere between 10 and 25% but I do know that as much as half of organizations will receive upwards of a 100% lift in revenue. So it can be quite profitable for you if you decide to do it with a thoughtful approach.

Ryan Greene: 01:53
Yeah. I think what we've seen, what I've seen as I work with different clients is the trend in the past, call it five- ish years, is that it's getting more and more expensive to acquire customers. And actually the research is, it's growing at 25% a year. And really, there's a monopoly on where you're acquiring those customers from, especially in the digital space, there's basically two or three sources for that, and they basically have a monopoly on that channel, and you're going have to pay whatever you need to pay. So then it becomes, okay, if it costs more to get these customers, how much value are you getting out of them? Right? And this is the gap, right? The gap between the cost of a new customer and the LTV, lifetime value of that customer. And if your cost is going higher than your LTV, there's a gap, right?

Ryan Greene: 02:40
And that is the experience gap. And I think today the reason we're having this conversation is, I think it's basically agreed upon that personalization is the way to fill that gap, right? Better experiences, whether you call it experiences, marketing, communication, interactions, is how do you make each one of those relevant to your customers, so that the ones that you do bring on and they're valuable, how do they stay longer, and how do they drive it up, the monetary value that you said, and have that kind of beneficial relationship both on the customer side and on the business side?

James Myers: 03:20
So, that makes a lot of sense. Now what we've also learned, as serving both B2C and B2B organizations, we know that personalization is different for different organizations. So, right? Whether I'm a retailer that specializes in speed and low-friction in the shopping experience versus a retailer that works in kind of a very hands-on, personalized touch experience, we see different approaches. But I know that, I know, Ryan, you've worked with a number of our clients on that kind of distinction, and you have an important perspective on that.

Ryan Greene: 03:59
Yeah. I think that's why this whole topic is a bit overused, and why people are kind of saturated and kind of exhausted by everyone coming in and saying "you need to do this better and better and better". And it's "what is this", right? Like what does that actually mean? And it's different for every customer. And so when we say, "we're here to help you", or "we help our clients personalize these experiences", the first question that the...or the first thing before that question should always be, what does that mean to me, to my business, and to my customers? And what has been proven, when you do these studies, is personalization is really how are you adding value to each customer experience that makes that customer's life easier? Whatever job, whatever goal, whatever task they're trying to achieve? And every customer, wherever they're at with your brand, is in that different space. But every business should know what those things are, right? And they should know, as a customer goes through their life cycle within that brand, what are the various pain points or friction or inefficiencies or confusion that their customers are facing, and then how do they take those and then build personalization capabilities and strategies around that to make those things easier? Because those are really the things that,when you ask a customer what do they want from their brands, when they talk about personalization, it's those things. Are you helping them through tailored and valuable interactions?

James Myers: 05:36
Yeah, yeah. No, I completely agree. And the tricky part here is, that while each customer is different, you need to be able to tailor to those specific customer needs and make their life easier. And so, if we were to use two examples, I might say some of our clients today, right? Pick a high-end retailer and then maybe pick a more kind of transactional retailer. And the desire of that consumer on the low-end kind of, more-frequent-purchase retailers is they want speed, they want the ability to be recommended their next best product. Whereas, at the higher-end retailer, it's quite a different expectation, right? They're looking for that, that kind of ushering and that kind of completing the image or completing the clothing set. What this set, like the occasion that most might be the best for, the kind of like, being almost a personal assistant, as opposed to the transactional retailer, which is really just trying to make your life as quick and easy as possible.

Ryan Greene: 06:47
Yeah. I mean, to put that into tangible examples, I think, what is the customer expectation when they interact with your brand? If you're Amazon, it's about convenience, speed, and price, right? Those are the things that they're going to, they invest all their money in, and if you're going to try to compete on those criteria, you're going to lose, right? You're not going to beat Amazon or a Walmart on any of those criteria, so don't try. Right? You have to try to compete on the experience side. So, the experience side, if you think about, with the customer that we have, Intermix, which is part of the Gap brand, they're all about, building that personal shopping experience. Their customers expect Intermix associates in the brand to know who they are, what their styles are, and then help them curate a wardrobe around that, and let them know when new interesting product is in the store, so that they can kind of be ahead of the curve. And they don't want to, or they don't expect to have to do that on their own. So they expect the brand to really bring that to the table for them.

Ryan Greene: 07:46
And for them, that's personalization, right? For Amazon, it's about how do we get you through this transaction as quickly as possible and get you your product as conveniently as possible. So, really, the first step before we get into any of these topics that we're going to get into today, is to answer those questions as a brand for yourself. Like what are those, what is the expectation? What is your customers trying to do? And how do you meet those, and go above and beyond those? And then start to build your strategy around what are the types of experiences or interactions that need to be in place to deliver on those. .

Ryan Greene: 08:17
So now that we've answered that, the next topic is about some of the, of what's in the market, the technology around personalization, right. Mark me up if you'd like, on the scoreboard.

Ryan Greene: 08:33
But it is probably the most overused term out there, which makes it a confusing space both for the buyer but also for a seller, a vendor like ourselves. It's like, how do you cut through the noise and really explain what do you, what is your approach to personalization? So, in this conversation, this section, let's talk about what's real and what's hype. When we're talking about, if I'm a buyer and I'm out there looking at, how to do personalization. Say I know what those experiences are that I want to create, but like how do I start to evaluate the different solutions, technology services, out there to build the infrastructure or the capabilities within my organization to actually deliver on those?

James Myers: 09:21
Great questions. And I think we'll consider framing around two topics. The first is, the amount of data that you need. And the second is, the speed of that data or the speed of your delivery of the experience. And so, the first one, if you think about the amount of data that you might need, from my perspective, it's that you need as much data from your customer interactions to their maybe purchases, demographic information, website interactions. You need as much of that data as necessary. And that's just to get it going. From there after, it's about increasing the amount of lift, incrementally, with additional data sources. But there's a common term in the marketplace called a 360-degree view of your customer, and while at a high level, I agree with that approach, where I don't agree, is that you need to start by getting all your data in one place just to get the ball rolling.

Ryan Greene: 10:20
Right? Yeah, no, yeah. I mean there's, there's many people that said it better than this, but - a lot of lot of brands are data-rich/insights-poor. They have rounding in the amount of data that they have, which for the last few years, it's all been, everyone's been talking about just get more data and the answers will come. Bring the data, the answers will come. But, I think most brands have found that that's actually not the case. It's not just get the data in there, and then stitch it all together. It's like, what can you do with that? How do you get information, intelligence, and insights out of that? And then, most importantly, how do you act on that, and how do you take that data and turn it into insights, and then turn it into experiences? Because that's really where the value is created, is at the experience moment, and whether or not you're impacting that customer's interaction with your brand, that then determines if they're going to want to come back.

Ryan Greene: 11:19
And so, I think definitely, starting with the data, what's real versus what's hype? I think definitely thinking about the perfect customer record probably doesn't exist. And there's no black magic to it. So, if you're talking about, I'm looking for a vendor or a solution that's going get me that perfect view of my customer, going to get all that data in one place, it just doesn't exist at this point. And so, you have to be okay with good enough, with the ability to build on that downstream. Because you really need to get value out of data right now, before you do overhaul your entire data.

James Myers: 11:57
The reality is that, just getting your data in one place is going to cost money. Whether that's new technology investments, or technology plus people-time, particularly IT, who often has to allocate their time to a particular department or a task, you're going to be asking for some amount of cost. And so, for that reason, you're going to need to show return on it. And the best way to do it is by taking RG's (Ryan Greene's) approach, which is: identify what your customers are expecting you to improve in their experience. So, where's the friction today, and where can we alleviate it to the personalization? And identifying which resources are needed to pull the data points that will actually improve that experience and provide a helpful outcome.

Ryan Greene: 12:45
Maybe talk about real time. And then we'll talk about on both ends, but data generated and ingested in real time. What's your feeling about the real timeness of data, versus the, not completely to separate, but I think as important, but maybe not considered as much as the agility, or flexibility, of the data that's coming into your platform? Because there's three things. There's volume, there's velocity, and there's variability of data. Right? And we talked about volume, like, you already have enough. What velocity is about, the real time is, is the variability. Veracity is the other one. You can think about that as the 360. But the variability is you have all these different data sets, and you don't necessarily know what you want to do with them at day one, but you know that maybe in the future there may be a use for them. So maybe talk about philosophy versus [inaudible].

James Myers: 13:41
I'll be a little controversial. It may go against what some others typically think. But, I would say that it's more important to have agility with your data than it is to be real-time through data. And the reason why I take that approach is because none of us have a crystal ball. And so, we can't figure out where we're going to need to create real time integrations, or real-time activations, on day one. And we don't have that type of foresight. And so, instead, what I think is more important, is to have the ability to access and manipulate your data as you see fit, which means being independent from technical resources, independent from IT, independent from going to the bakery and taking that ticket and saying, I'll sit here and wait while you have time to service me. So, I think it's more valuable to be able to ideate and iterate with the accessible agility to do that today, as opposed to starting from day one and saying, I got to have all this in real time. What do you think?

Ryan Greene: 14:44
I mean, it's never an either or an or, right? It's, you want data to be...the way that we always think about it's use case specific, right? You only need your data to be, you only need as much data or as real-time data as you need for the use case you're trying to solve. Right? But you don't know the use cases you're trying to solve in the future. You know what you're trying to solve today. So, build for that, which makes that variability piece very, very valuable, because it's those future use cases that you don't yet know today, but you know there will be there. And so, if you spend all your time on volume and velocity, you may figure out and get in place what you need today, but then, how quickly did those requirements and these use cases change? And then, you're six months in, and you're like, well, I've got to build this new thing.

James Myers: 15:30
Let me ask you this question, right? How many times is that here in previous jobs have you been sitting at your desk, or you're sitting in a meeting with others, and you have this new idea, and you go, I know this idea is going to work. Maybe it's not a 100% probability, but you know, 95%, you know the idea is going to work. But then, your next step is, oh, wait. In order to make that idea a reality, I need to go get this data, tie up that resource, integrate these systems, and then by that point you're sometimes montths down the road. Does that happen to you?

Ryan Greene: 16:02
All the time. Did you think about the cost-benefit relationship? If it's super expensive to try something new, you're going to try less things. Which then decreases your innovation, decreases the experimentation, and decreases the amount of things you can put out there to see what works. And so, you're kind of stuck in this static place where you, some of your data is real time, and you have a lot of data and maybe you have a 360 view, but you're kind of stuck to these certain number of use cases.

James Myers: 16:31
It's why, I will say, that just before we close out on real time, if there's one thing that I do require, I don't want to see your display ads for days after I make my purchase. So, if it real time, to me, in my mind, is less than a second, near real time can be hours or half a day.

Ryan Greene: 16:52
It always comes back to the use case, right? It's real time for that use case. Real time when you're clicking on something in the website and you want to personalize whatever comes up in your page, that's milliseconds. So, if it's, I bought something in-store and I don't want to see the same thing when I go home on the internet. Maybe that's ours. But it all, it's all relevant as they.

James Myers: 17:13
Good point.

Ryan Greene: 17:15
Okay. Moving on. Data into intelligence, right? Analytics, AI, very hot topic right now. A lot of different claims out there, a lot of different promises being made. What's real?

James Myers: 17:34
Well, actually, I think that from an AI perspective, the word that bothers me the most, is when organizations claim to have AI or claim to have machine learning. It's unfortunately, it's typically a marketer, or it's typically someone on the business side who's saying, I want to use this buzzword. And, and it's not like you shouldn't have that as your goal, don't get me wrong. But the term AI, and the term machine learning, to me, are very elevated definitions, and they should be true to what their original definitions were, which is: machine learning is a feedback loop approach, wherein the algorithm is becoming more precise, and more accurate as time goes on because of that feedback loop.

James Myers: 18:17
And so where I would ask a question is, if I don't have that type of feedback loop, so call it true machine learning, can I still get value from an algorithm that's just playing out of the box or do I need that?

Ryan Greene: 18:35
I mean, you don't need it. It's more valuable, right? Because, you're just talking about elevated capabilities. But, again, it's all into, for what purpose? I think one of the big misconceptions when you talk to brands, is that AI should be both asking and answering your questions. Right?

Ryan Greene: 18:54
So, that's one of the biggest misconceptions is that, by putting AI on top of your data, they're not going to run your business for you. They're not going to understand all the nuances of your customers, understand your business strategy, your business objectives, your business constraints, and build this, always-on running marketing machine behind the scenes, right? It is, a tool that is, needs to be pointed in a very specific direction, and needs to be tailored to very specific criteria. And so, sure, you can, basically what you're taking is human intuition and human decision-making, and you're scaling it up with machine learning for better or for worse. You're taking the ability to be human, to look across a certain amount of signals and make their best guess, and then expanding that to millions of signals and always, you know, optimizing, and getting to better answers over and over across a lot bigger base of data.

Ryan Greene: 19:59
So, I think the idea of "out of the box" doesn't exist, because "out of the box" assumes that it's tailored to your business needs and to your objectives, which just can't be, if it's out of the box. And the feedback loop, also defined, is determined on, do you have the amount of data that is needed to find these patterns and these signals? And then, this goes back to your data, do you have the right data, the right volume, to be able to find these patterns? And are you feeding this back in, and then rerunning the algorithm to get to a better answer? Or are you just putting rules against these things? Like, if/then statements? That's okay. It's not bad. But it's not machine learning.

James Myers: 20:43
At the end of the day, is that better than nothing? Yeah. Will you get an incremental lift from that? Yes. Will you get even more if you go with a machine learning approach? Absolutely. But to point Artie's point, I think it's important that that restate it, which is, the technology is not going to drive your business strategy at best we observe in the marketplace today is that machine learning and artificial intelligence can provide some level of prescriptive insights, but it's not going to tell you, "Oh, your next quarter you should have invested in this channel or that goal, but this quantity", and without a human being able to say, "Oh, wait a second, we know our business a little bit more than the machine does and we're going to override that type of approach".

Ryan Greene: 21:34
Yeah, and I think the other part with artifcial intelligence is the iterative aspect of improving over time. It's not, you can't, again, you can't just feed a machine to say "optimize for this outcome". You also have humans in a loop that are looking at other aspects that weren't necessarily planned for, or assumed, or other things that you can't necessarily program into a machine and allow them to take those learnings and change things on the fly, and see if that improves the business, because really, where things improve over time, and you can say definitively "this is better than this", is through experiments. And it's not through, this, again, may be controversial, but it's not through multi-variate attribution modeling, right? It's still a little bit of black magic, experimentation test and learning, looking for statistical significance is where really you can say for sure this is better than that.

James Myers: 22:28
I agree. So, let's move on to orchestration or delivery. We think about some of the hype in the marketplace today is that between federated decision making and centralized decision making. And we hear about it a lot and, and it should, because the reality is that consumers are engaging in more and more channels than they ever had before. And they're looking for that personalized experience. So the question that organizations come to us with is, "should I have a centralized brain or should I have a, a federated set of brains?". What's your perspective on that?

James Myers: 23:12
I like the term, there's too many cooks in the kitchen. It reminds me of that. Actually, you can probably think about it as like a restaurant, right? There has to be a head chef, who's designing the menu, who's taking the ingredients and coming up with, this is the, this is what I want the experience to feel like. But then you have to have these, the line cooks actually in there and making decisions on the fly, on their particular meals. Are there particular dishes that can't be accounted for by the head chef? Right? So that can be a way to think about it. Right? You probably definitely should have a centralized way to think about your customers, regardless of channel, regardless of stage. And kind of set these rules and strategies and intelligence, and then push that into those channels, and allow them to do what they're really good at, which is delivery and kind of the contextual setting, as well. But it's all got to be based on that same kind of coherent strategy and decision-making that's done by the central kind of known.

James Myers: 24:23
Yeah, and actually, I liked the analogy of, so let me pretend that there's one central chef, and the chef says, "I know the consumer types, they're going to come to my restaurant and I know the types of meals that each one of them are going to like". And so, that central chef has that wisdom, because they've scaled and they processed all the customers that have ever shopped there before. But, what they're missing is the on-the-fly context. And when Ryan comes into the restaurant, and actually says. "I might look like Customer A, or I used to shop like Customer A, but now I'm feeling a little different today. I want to try something new". And so, what you would talk to the waiter or waitress about is new intent. And so, that waiter or waitress has the ability to take the information from you as the customer, and combine it with the recommendations that typically come from the chef, and say, "here's the optimized product or meal or drink that would be best for you, because I know your current intent".

Ryan Greene: 25:28
Right. But you don't want every waiter or waitress, or a cook, making the entire decision on the fly of "and this is what I think this person should have". And then, the next person next to them or the table next to them is getting a completely different experience, and there is no coherent field, un the entire experience, the entire restaurant. So, you have to have that one centralized place to make those decisions.

James Myers: 25:51
Yep. Completely agree.

Ryan Greene: 25:53
All right. We have a couple minutes left. So, why don't we close out with kind of trends and where do we see this all going? And maybe, we keep it across those the same dimensions, data intelligence and delivery, and just kind of riff on what do we think, where do we think the market's going and say the next 2-3 years.

James Myers: 26:10
Well, you know,. from a data perspective - I'll go first real quick - I think the common trend that we're hearing is that organizations are looking for a technology stack that enables a best of breed approach. It's that kind of a "smart hubs, dumb spokes" approach that my own team, had coinedwhen we were at Gardner. And the thinking here is that organizations are getting more agility and more effectiveness out of those types of stacks than they be from a a one-size-fits-all integrated suite, a la Oracle, Adobe, Salesforce. And actually, there was some good research that Gartner published just a couple of weeks ago about this very topic. But, what you're seeing is that organizations are actually changing their marketing technology stacks upwards of once a quarter, if not faster. And that the percentage of organizations that are doing that is 68%. So, 68% of changing your stack every quarter. If not more often.

Ryan Greene: 27:16
That's more than half.

James Myers: 27:17
Well, yeah. So, the reason why is because there are more channels that consumers are engaging with, and there's a higher expectation that they want personalized experiences to them. So, you need to be able to serve them, and adapt and innovate quickly. So that's kind of my data on...do you have one?

Ryan Greene: 27:36
Yeah, I think, a couple of ideas. A couple of thoughts on data, because this migration to the cloud is still happening, but I think that is for sure the trend in the market, and even the ones who are kind of the late movers - banks, more highly regulated businesses - they're all going to go to the cloud, right? The on-prem -

James Myers: 27:55
- the government version -

Ryan Greene: 27:57
Exactly. The on-prem versions just isn't going to be an option anymore. So, you have businesses moving to cloud, which is an opportunity for them as they change their data infrastructure, to change the rest of the marketing ecosystem to be able to keep up with the modernization of the data infrastructure. I think that's definitely a trend that we see. And then, what I see is, as you move your data in the cloud, putting more purpose-built data analytical engines on top of that, to really be very specific and very focused on the workloads and the jobs that are trying to be done, which means you can then be very specific about the types of data, the volume of data, the velocity of data that you want to bring into each of those systems, and not have to have a "Big Bang" project where you're saying, I have to bring everything all at once. And then I have to figure out how to use those across many different functions, right? Marketing, finance, servicing, operations, there's a bunch of different, very different needs and requirements and constraints across each of those functions. If you try to look for a one-size-fits-all approach, both on the data side, or, as you said, on the delivery side, you're kind of going to get the worst of all worlds.

James Myers: 29:06
Yeah, you're right. And then we could go on for many, many more drinks on that topic. But, I definitely agree with you. No one solution will solve for the entire enterprise. And it's really important that if you're not aligning the characteristics or the strengths of a given technology to the required use cases. A quick example.There's are some big data platforms that are built for the enterprise, and they do a good job of being a kind of a generic big database. But they don't serve the purpose of marketing or the customer or that type of quick computing, real-time response, big data processing. And so, you've got to be careful. It's not just like, you go out and you hear words like "spark", "snowflake", and the list goes on and on and go, "oh, that's the one I definitely need,. because I hear it every time that I walk down the street". You've go to be more deliberate than that. So, good point.

Ryan Greene: 30:13
I think the last one, just to close out, is the optionality, right? Which is, again, not trying to get everything all in one kind of solution or portfolio or cloud, but having the ability to swap out pieces as needed. And having capabilities and solutions and technology that allows you to do that. Which gives you the ability to be agile on your data, on your use cases, on your strategy. Okay. So, thank you all for joining. How did we do on the scores?

James Myers: 30:45
Yeah, probably not true. I think, I think you missed me a couple of times. I appreciate the generosity. It's been a lot of fun. Yeah. I hope you enjoyed it as much as we did. And, we'll be doing even more of these. I think there's plenty of beer in our internal bar. I know two people that like to do some taste testing from time to time.

Ryan Greene: 31:11
Here we go Okay. Until next time. Thanks, all.

James Myers: 31:14
All right.

Find out more about how Action IQ can help you.