Cracking the Code of Customer Experience with Dave Edelman

0:00Hey, welcome to the CMO Whisper Show. I'm your host, Steve Olensky. Part marketing practitioner, part ad agency veteran, part journalist. I was a writer for Forbes for 10 years. I've had so many insightful conversations over the years with business leaders, to athletes, to celebrities, to, of course, CMOs. The only difference now is instead of sharing those insights through written form, I'm doing it this way. My guest this week is back for another episode adventure. He was one of my first guests on the CMO Whisperer. His name is Dave Edelman, and he's a longtime thought leader in marketing and customer experience with over, ready for this, 1 million LinkedIn followers.

0:43Over 30 years, he's led strategy and digital marketing practices at BCG, Digitas, and McKinsey, and he was also CMO of Aetna. He now is an executive advisor and senior fellow at Harvard Business School, working with enterprise-scale companies and AI startups. His new book, Personalized Customer Strategy in the Age of AI, hit number four on USA Today's best sellers list. And for kicks and giggles, he was just named an AI trailblazer by Adweek. Dave Edelman, my friend, my confidant, my mentor, you name it. We go way back. Welcome back, my friend. Thanks, Dave. It's great to see you again and be here.

1:22It is such a blast. We follow each other on LinkedIn. We text. We check in. But I figured, you know what? It's been time. You were literally, I think, my second or third guest on the show. So you're a two-timer. We'll get to that Saturday Night Live five times. Right? So we'll have something to shoot for. Okay. I want this episode to focus squarely and solely on the book because I am fascinated, and I really want the listeners and everybody to hear your take and on the power of personalization and everything that goes with that. And you call them and you and your co-author actually give a shout out to your co-author.

2:03Sure. So my co-author is Mark Abraham, who runs the marketing practice globally for BCG. He's been a terrific comrade in getting this book done. And we've done speeches together as well. And it's been great. I'll have to. I definitely have to get him on as well. Yeah, you should. OK, so you've both identified what you call the five promises of personalization. It seems like when I saw them, like that should be on granite or something like the Dead Sea Scrolls. So what I'm going to do, I want to break them down one by one so everybody can hear them and have conversations on each and really get into.

2:43And you're so, I think, uniquely qualified to speak on these, having not only been in the consulting, but a CMO of a monster major brand at the same time. So really interested to get into this. So the first promise of personalization you both, you and your co-author identified, is called empower me. Understand each customer's needs and how to best meet them. First question I have when it comes to this is how do you ensure your team, when you have a team, understands the truly evolving needs of each customer that go way beyond the surface level data? Yeah, sure. So there's two levels at which you can think about empower me.

3:23One is actually at a higher level, which is not just thinking about pure marketing, but overall, thinking about the customer experience your brand delivers and how by using customer information, you can take that customer experience to a whole new level. I mean, look at brands we use all the time, Spotify for music, Uber, Amazon. The reason you're constantly coming back to them is because they know you and they use that for not just marketing, but to actually help your experience. So there's an empower me that's at the higher level in terms of designing the customer experience in the first place.

4:07And then there's a more micro level of understanding in a particular moment where you are and what you might need. So, for example, one of the companies that I've worked with, one of the big telecom companies, has used data to understand all kinds of issues that consumers face as they're going through their cable TV experiences. And they have desperately been trying to upgrade their level of responsiveness and how fast and complete they can make that experience. So by using data, they can totally understand right now that you just went from the app to call center. Okay. You went from the app to the call center.

4:58That means something went wrong on the app. So then let's go back and actually see on the app. What were you trying to do? Oh, you were trying to pay a bill. Oh, you were on Apple's iOS phone. Oh, we just did an upgrade. There must be a bug. And so you're tracking an individual customer back and you're using that data to think, so now I know there was a bug. You had a problem. You're in customer service. So I can send a message right to that rep to explain to you that there was a problem with the software. Try again in a half hour. It'll be resolved. And so you're constantly trying to understand the context of individual customers through various kinds of data.

5:46And the challenge, which gets to the next one, Nomi, which we'll get into, is what do you want to understand about that customer in order to empower them to do something? So just another very simple example, when you, Cisco, the food delivery company, they have an app for people who buy food for restaurants. Now, if you're buying food for a restaurant, you don't want to spend a lot of time buying food. You have narrow margins and you got a lot of other things that you want to get done. When you open that Cisco app, it knows who you are, where you are, your restaurant, your menu, your price points, whether you buy in bulk or on spec.

6:30And it can actually prioritize offerings for you specifically of food and things that are appropriate for your restaurant and the way you buy. That's empowering you. You don't have to start going hunting and everything. it knows who you are and makes it easier. That's what we mean by empowering. It's that they are combining both those things. They're thinking through an experience that works for you as an individual. And then as a micro level, they're understanding the specifics that you're calling in right now. Maybe it's a certain time of year. They know already that you just bought a big bulk of something else.

7:09So you may not need any more of that. And all of that intelligence is being use to make your buying experience easier. Exactly. And I think, you know, I might have got a little ahead of myself when I was introducing these five promises, you know, going forward, I'll, I'll state them, let you speak on and then get into the questions. So I was just so excited to get into this one because all five of these are so relevant, so germane to everybody. And they say better be in marketing. So let me get a little more granular now that we've explained what empower me means, right? So in your, in your experience, what are some tools and strategies that have been most effective?

7:45You just get a great example of Cisco. What are some tool strategies that you would recommend? Yeah, so the whole thing is about taking a philosophy of saying where are customers facing compromises in the way they get things done that we can actually address by using customer information about them in a helpful and appropriate way. So it starts, there's a mindset that says you want to help solve customer problems. So an example of that is when I was chief marketing officer at Aetna, one of the problems is people don't understand their health plans. That's a challenge. That's not a good thing. It's just inherently hard.

8:30And so we needed to step back and really rethink from the customer's perspective, what did they need? Not just from a compliance perspective. And so instead of just throwing paper at people, we eventually realized that we could come up with a more engaging way to do onboarding. So we came up, we found a company called Sunday Sky that could do personalized videos. And every new Aetna member got a personalized video explaining the specifics of their health plan to them. That was powerful. Calls to the call center went way down. People started opening up our emails more. We realized we cared. We even told people about urgent care centers that were close to where they lived.

9:17So we were able to reduce unnecessary visits to emergency rooms because people could go to the urgent care centers. So it's a philosophy of saying, where are our customers in what they want to get done? Where are they stumbling? Yeah. So let me play devil's advocate for one quick second here. Yeah. I'm a marketer. I'm listening to this going, that's great, Dave, but I don't work at a company the size of that. Now, I don't have the budget to start, right? A video start of Sunday Sky. What do I do? Well, first of all, a tool like Sunday Sky, some of these tools are actually not expensive at all.

9:55It does not require elaborate technology. And now with generative AI, which we actually didn't have at the time when we did this, it's even easier to do. So some of these tools are actually a lot less expensive and simpler than you may think if you investigate them. The other thing, though, is that actually small companies may have more of an advantage than a big company. So when we looked at who were leaders in personalization versus those who were struggling and thinking about it but having trouble, the biggest issue between the leaders and the laggards was the way they operated. It was not whether they had the data.

10:40It was not whether they had the technology. A lot of the common tools out there, there are tools available, and a lot of companies may already have a CRM software that has some capabilities, but you're just not using them. And the main thing we found was the way companies were working, their mindset, how they thought about creating value for their customers, and how they thought about marketing not as a way to just push out broad campaigns, but as a way to think much more about individuals and have touches that are appropriate for an individual and line up their CRM program, their trigger program, their retargeting program.

11:25programs in ways that are just simply more intelligent, smarter, more strategic than just simply bombarding. Exactly. Okay. I want to make sure we get to all five of these promises. Let me move on to number two. And I'm going to read it verbatim. Know me. Win customers' trust and permission to use their data to improve their experience. Talk about that one. Right. So in order to deliver experiences based on customer information, you got to have customer information. And you got to do so in a way that consumers are going to be comfortable with. And we actually did research on thousands of consumers in the U.S.

12:05and in other countries about their attitudes towards companies using customer data. And it was fascinating. About 75 % in the U.S., it varied a bit by country, but in the U.S., 75 % of customers were comfortable with businesses having information about them and using it to give them special value. Now, the flip side of that, though, is two-thirds of those customers said they've stopped doing business with companies who use their data inappropriately. So there's a bar here, and that bar may vary by different customers, something you're going to have to test. But we need to start with what's the data you're naturally collecting in the course of doing business, what we call first-party data.

12:52And often companies have all the marketing data. They may have sales data. They may have service data, product usage data. But they're all in separate databases. They're not brought together in any integrated way. And one of the things we talk about in the book now is that there are AI tools. One of them, for example, is from a company called Narrative.ai. And these allow you to integrate data on the fly in ways that you just couldn't do before. So you can use this software to look at one data set and it understands the schema, look at a second data set, understand the schema, and then it can write the code to integrate those two into a third receptacle.

13:36So some companies use this, for example, as a front end with Snowflake to pull in data from multiple systems into a CDP. Now, that was not possible before. And now it can happen on the fly. And you can bring together the knowledge you have increasingly and faster, faster, faster to be able to use that to understand more about an individual customer. So that's one side of things. The other side that is very underrated is what I call zero party data, which is just when you ask people questions. It's shocking to me how few marketers ask me questions about what I'm interested in. They constantly ask me questions on my satisfaction of this experience, that experience.

14:24I'm like overwhelmed with all these satisfaction surveys, but they don't ask me what I want. And in the book, we quote Peggy Rowe, who is the chief experience officer at Marriott. And they have an explicit strategy to ask Bonvoy loyalty program members questions like you're a frequent business traveler. Where do you want to go on vacation with your family? What amenities would you like in your room? What are the things you might want to do in town? And you can actually see what they have about you. You can edit that. They use that to add value. And it's, you know, that people tell you, why aren't we doing more of that as marketers?

15:05And it's open-ended questions. Yeah. Right? It's open-ended questions. And the AI capabilities now can turn that into real data. But you have to have a philosophy that says you want to do that. and too many marketers aren't thinking about that. I know. It fascinates me just from a human nature perspective. Why would you not want to know? Yes, you're going to have some stuff that's not great and not positive, but okay, that's how you learn. Yeah. So how do you, saying on this permission and data, how do you find that right balance between personalization and privacy? It's a fine line, and one person's comfort zone is another person's creepiness, and that's why you've got to test it.

15:47You've got to test and test and test and you test on small samples and you see how people react and then you grow. I mean, that's what we did when I was at Aetna. We were constantly testing new things, often with our own employees starting there as customers and seeing how our own employees react first, which is a safer zone. But then using what you learn to keep tweaking and figuring out where things are going to be comfortable. But ask, just simply asking people, you know, for example, we realize, you know, people buying Medicare, it's complicated. So we just we came up with five simple questions that to ask people about, do they travel much?

16:30You know, do they do they feel a need for care for a chronic condition? Just things like that. And then based on that information, we could start thinking about what's the right plan for them. And then once they're on the plan, how to help them get the most out of it. And they're voluntarily giving you information. And there's just such an opportunity there to personalize based on that. Okay, next promise of personalization. Reach me. Reach out to the right customer in the right channel at the right time. Sounds so simple and so direct, but we both know it's not. It's not because so many things and the way marketers run campaigns conspire against this.

17:14And let me give you a simple example. So a year ago, I moved from Lexington, Massachusetts to Cambridge, and I upgraded my coffee maker to a Nespresso. I love espresso drinks. In fact, I'm holding one right now here. So the Nespresso comes with 100 pods for making coffee. Three days after I get my Nespresso and my 100 pods, I'm getting bombarded by Nespresso with discount offers to buy more pods. How much coffee? Three days. How much coffee do they think I drink? And if they thought we were like a small business, they have my address. They just shipped me all this. They can see that we're a resonance.

17:58And so that was ridiculous. And then it got worse because the following week they started pitching me to buy more machines. And so what clearly happened is they sold me a machine and I'm just now on a list. And they just bombard me without really thinking about the context. What they could have done three days later is ask me, how do I like my machine? Do I know that I can make cold as well as hot drinks? What flavors do I like? Would I like to go on a coffee tour of the world and try different flavors and learn about them? Go on auto replenish. They didn't do any of that. And so I unsubscribed.

18:41And when I need pods, I go and get them. Yeah. But that's not reaching people appropriately. And it's not thinking about the recipient from the point of view of them interacting with your brand. So as far as reaching, it's, you know, the operative word obviously is right. So it's right customer, right channel, right time. So is it as simple operative word as looking at the data and says, well, this group of customers prefer this channel and so on? Is it that? Well, yes, but, yes, and. So from a channel perspective, yes, but there's also different situations when a channel could matter, such as you're on a street near one of the stores and you like a particular store and we know you're an avid loyalty member of a particular store.

19:33So you're out on the street near it. We're going to pop you an app notification or text saying that certain thing is on sale that you might be interested in the store. Whereas normally we might send you something through email. Your contextual situation at the moment is such that this is a channel that's going to be appropriate. Now, that's one side. So but then there's also timing, which is thinking about the orchestration of messages. Are you just going to bombard somebody or are you going to really think about less being more so that if I don't bombard you all the time, you're going to actually pay attention?

20:14So another thing when I came in as CMO at Aetna, I was one of the customers now of Aetna because we have our own health insurance. And I started looking at all the stuff my own company and my organization was sending me. And I said, this is crazy. It's out of control. And so I had us do a scatterplot of how much were we sending to people against how much of that is opened. And it was very clear that if we send more than four things in a month, people start tuning out. Unsubscribes go up. Open rates plummet. So let's be smarter. Let's pull back. And let's think about the most important things to send and focus on those.

20:54And sure enough, once we did that, we got a lot more engagement. Yeah. On the right channel, right time. You know, again, not to sound Captain Obvious, but that's the point. Don't overthink it. Okay. The next show me, tailor unique content to be relevant to each customer enabled by generative AI. Now, this to me is very, I want to break this down if I can. The first word that jumped out at me when I read this one was relevant. Again, very overused word, but it's so important that you're going to tailor this content. but to be relevant. But now you're enabling it by Gen AI. So elaborate on that one.

21:34Yeah, sure. So earlier I mentioned the Sunday Sky example of giving you a personalized video based on what you have bought from Aetna and who you are as an individual. So that's, you know, a personalized video right there. Give you another example, the town where I used to live in had an incentive for people to get solar panels on their homes. And if you bought a solar panel, you get a reduction in your property taxes. And that led to just an avalanche of marketing, everybody trying to pitch solar panels left, right, and center. But one pitch said very clearly, Edelman family, we've done the math, and we believe you can save over 20 % on your annual energy bill with and the company was called sungevity there's a personalized url here for you to click on that'll explain why so i open it i click on it and i get a google earth image of the roof of my house with solar panels super imposed on my roof a calculation next to it of how much energy those panels would generate based on the number of panels, the longitude and latitude of my house, and the east-south-west-north orientation of my roof.

22:54So that tells me how much energy. Then they went to Zillow, got the square footage of my house, and used that to estimate how much total energy we might use over the course of a year. So they have the numerator from Google Earth, the denominator from Zillow, it came to 21.3%. Then from that point on, everything just continued in that way in terms of completely personalized interactions, which we eventually did. So that's showing, showing me something personalized. Relevant. Very relevant. They took something that was complex, made it easy to understand. And they certainly got me emotionally by seeing that image.

23:39And it didn't cross the creepy line because it's all public information. It's all public information. It's also relevant to the fact that my town has a discount on solar panels. So it wasn't creepy. Sometimes when I've told that story, people do say, oh, that would be creepy. But you know, it's out there. It is. Your square footage, like you said, Zillow, it's all public. It's all public. Yeah. That's interesting. That's a great example. I love that. So, okay, devil's advocate here or not even devil's advocate, a marketer is listening. What's the key then to delivering that that feels personal and relevant, but not automated?

24:17Yeah, so some of it is automated and it depends on the nature of what it is that interaction. The Sunday Sky video, what's longevity sent were automated, but they were designed to be clear, understandable. When we did the Sunday Sky videos, we actually tested, oh, probably easily 15 iterations of different video styles, fonts, ways of doing it to figure out the best way of connecting with people. So you have to test. And again, I often come back to test and learn, which we'll get into in the very last promise. But that's absolutely critical of understanding how people are going to react to different things.

25:03And the human element is all about the design in the first of the concept of what you want that experience to be for the customer. And then making sure the design that comes out not only reflects that at the high level, but brings out your brand. That's the art of creative. And that's why we're still going to need creatives. They may be enabled by AI. We may need fewer people to bang out content per se, but we need the design. We need the strategy. We need the emotion. Otherwise, it's all going to look like robotic crap. Exactly. And then they're done. Then you're out. Okay. We're up to the last of the five promises of personalization.

25:48Before we get to that, let me recap here. Number. So we have empower me, know me, reach me, show me. And then the last one is the late me, which you describe as design new ways of working and ensure continuous improvement. So a customer's experience feels magical. Flush that one out for us. Sure. So think about your interactions with all of those digital brands, ranging from Spotify, Amazon, Uber. The more they learn about you, the better your experience, the more personalized that's going to be. And it's happening at two levels. One is they're getting more information from you about your interactions.

26:28Marriott is getting more information about you as you do hotel stays and as you answer their questions. So they're getting more and more information about you personally that allows them to do things. But they're also, in the broad sense, learning more in general from all the interactions that they're having. People like you who may be doing things that are ahead of where you are and they start seeing how those people react. And so the ability to test at scale and learn at scale becomes absolutely critical. And as I mentioned earlier, the biggest difference between the leaders and the laggards in personalization is the way they operate.

27:14is thinking about the way you go to market so that you can get things out quickly, measure what happens, and then use that to iterate, innovate, or optimize on a constant basis. AI needs to be fed. It needs constant flow of data to get smarter. And the more you can go out to market, the more interactions you can have, the more you can test, the more you're getting feedstock for your AI. And that's your feedstock from your interaction, not the same ones that other people are getting, other companies. So you can maintain a smarter advantage based on your own learning as the AI gets smarter, whether that's AI for content generation, for segmentation, for targeting, all of that gets smarter as you learn.

28:05So delight me happens as you constantly are out there moving in fast cycles. And there are things I've talked about through a big chunk of my career about agile marketing and using the principles from software development applied to marketing. And agile doesn't just mean fast. It doesn't just mean responsive. There's actually a discipline to it around smaller groups of people empowered to do things, working on two-week sprint cycles, stand-ups every day, a minimum of handoffs, and really taking advantage of the automation and capabilities at their disposal. That's what I spend a lot of time with my clients working on.

28:49A big chunk of my own advisory practice is helping clients get those kinds of engines going. But it's a different mindset than just 13 people taking three months to get a campaign out the door. Exactly. Big, big difference. And you have to accept that. So, okay, looking at the clock, Winding down. I don't think I was doing this the first time you were on, but I called the random five and it's five very random questions I ask each guest. And they're all meant to just have the audience get to know you a little better and they're all fun. And if you're game, here we go. All right. I just had another shot of espresso.