The Loyalty Puzzle: An Interview with Dr. Yuping Liu-Thompkins, Founder, Loyalty Science Lab

YL: Thank you.

SS: I want to start though with a little bit of a provocative conversation, uh, with the provocative Byron Sharp. And obviously as a lifelong marketer, you're familiar with him, the gadfly. Now, of course, his position all along has been that loyalty is largely a myth, that consumers are naturally polygamous. And then I saw this December interview with him in Ad Age where he goes on to stoke the fire by saying “fashionable things like investing in your existing customer base” - I love his use of the term ‘fashionable’ – “and hoping that you'll get lots more out of it turns out to be wrong.” So, you're a specialist in loyalty. Is he wrong? And is it just that he's looking at things at the wrong end of the telescope? Like, what's the issue here that you feel needs to be addressed when he makes these provocative statements? [9:47]

YL: Yeah, happily responding to that. I think there is being some conversations around that, right? Speaking of ‘fashionable,’ I think How Brands Grow”, uh, mentality has also been fashionable with some of the big brands too. I think from my perspective, I see this as, yes, it happens sometimes in some cases that is how it works, right? Particularly when we are looking at a very mature market where brands have fairly stabilized and it's also considered to be a fairly low involvement, high frequency type of product market. We probably will see that kind of pattern.

At the same time, it would be a mistake for any brands, small, big brands, to basically say that's all you should be targeting, that you should become, people should become aware of you, you should be reaching, you know, be physically available as much as possible. I don't necessarily think that's the same formula that works for everyone. Because when you really think about it, if that's the only way it works, the market will never change because bigger brands will always be better. They have bigger pockets, they can go to more people's mind and more stores and so forth.

But history has told us otherwise, right? We have seen plenty of examples of smaller brands disrupting the market, initially starting with a really core loyal group of fans and then from there maybe then they can work on reach and those aspects in order to really expand. But I think where they started from is not necessarily from the idea of let's just get ourselves into as many stores as possible and blanket the market without advertising as much as possible. Not everybody has necessarily started from that point.

SS: I think too, you know, it has something to do with just the idea of growth itself. And, there are different types of growth and there's easy growth and hard growth. If you're talking about chasing after lower frequency, you know, light users and making up somehow through that pathway, it seems to be an expensive one. Versus getting your next easiest dollar is going to come from an existing customer and most of the time 80, 85% of your revenue is coming from existing customers. So I've never got that argument. It seems to me it's a, it's of an old world view of how marketing and advertising works to drive revenue. Would you agree with that?

YL: Well, I think it depends, right? So one of the concerns about targeting the heavy customers is that when we're talking about growth, right, we're not talking about maintaining your existing revenue, we're talking about growing your revenue. If you're thinking from that perspective, there is a real possibility that your heavy users have already maxed out their usage. So there's always the ceiling effect, right?

And so from that perspective, if your industry or your customers are actually facing that situation, I only need so much X and if I've already devoted all of my share of wallet with your brand, then you, you can keep on talking to me. It's not going to make a significant difference to what I would purchase. And so, in those kind of contexts I would say yeah, maybe they're not the best target. What you wanted to do there is maintain your relationship with your loyal customers.

I think that on the really extreme end of things, light users, it also kind of depends on the mentality. Are they light users because they really don't have a need for your product? Are they light users because they really don't like your brand? Or are they a light user because they are barely aware of who you are, right? And so, each of those three categories you kind of try to target them, you're going to get very different results. And so, so I don't necessarily think it's a blanket kind of statement that can be made to say that's where you need to go. It really depends on the nature of the brand, the nature of the market.

And I will also mention that years ago when I first started doing loyalty-related research, I had actually studied the effect of loyalty programs on customers and how the reactions actually are different based on the usage level of the customers. So those studies, I think one of the insights coming out of that is that yes, with loyalty programs what we saw was the moderate users and the light users actually show the most behavioural increment from participating in the loyalty program. However, we also see restrictions where once the market becomes saturated with loyalty programs, you know, most competitors, they have one already, then what you're essentially looking at is number one, can people still be flexible in demand, right?

Taking travel industry for example, people can decide to take one more trip, one less trip, you know, assuming they can afford it. So in those kind of situations, even if the market is saturated, you can still have opportunity to grow by targeting some of these moderate and light users in that case. But in situations where you have very rigid demand, where it's a fairly set demand, then it becomes more or less a zero sum game. [15.16]

SS: Well, we'll come much later in this conversation around the concept of customer portfolio management and different strategies to apply to different segments for sure. But I do want to do a deep dive obviously into loyalty, broadly speaking, and certainly talk about programs and program optimization, et cetera. Measurement is another area we'll explore. Let's just start with your definition of customer loyalty. And you make the point that loyalty is not a monolithic concept. It means different things to different people. There are different, I would say, gradations of loyalty, if you will. What are those different segments or definitions as you see them?

YL: So I would say my view of loyalty is very much coming from a long history of, you know, both my own research as well as previous researchers exploring the construct with consumers to try to really understand what does loyalty really mean to a consumer and, or to a business, ah, for that fact. I sort of use the analogy that loyalty is kind of like an onion. You know, you peel one layer, there's another layer in it. The difference is that for different consumers, they might hold different types of loyalty towards your company.

So on the, on the most outer, you know, layer, when you, something that you can really observe is the behavioural loyalty, right? So from a business perspective it’s quite obvious. You know, if you are buying my product a lot, you're always coming back, you are repeatedly returning to my store, then I'm going to think of you as a loyal customer. And so we capture that through for example customer lifetime value, you know, whether you're used, you're using RFM models and so forth.

So that's one layer that more observed kind of behaviour from a purchase behaviour standpoint. But then if you go a little bit deeper into it, there's also the aspect of attitudinal loyalty. And so this is where you know, we talk about things as brand love. We talk about like the true loyalty. Some people use true loyalty to represent that kind of situation where people really just love your brand, right? They follow your brand, what you do, they advocate for it, they talk to other consumers about it. And so, so this, this layer of loyalty, I think it's very precious. I think a lot of the brands can potentially get to the point where they have the behavioural loyalty of consumers. But perhaps the moment a competitor comes with a slightly better offering, those consumers are going to be on the wall, not necessarily going to be loyal to you anymore. But I think that core loyalty inside of the consumers, they have you as the better brand, they connect better with you. I think that's the really true attitudinally loyal customers.

But then another layer of it is probably the very beginning part of forming a relationship with a brand is something we call cognitive loyalty. And so cognitive loyalty is more just about, I know this brand is better, right? So a customer can be loyal to Amazon because they just know that Amazon can deliver really fast. Or, you know, a bigger brand, I love Pentel pens because I know it really just writes better, right? And so, so some of them may not necessarily always involve emotional kind of loyalty because face it, not all products, I mean will be emotionally exhaustive. We're emotionally loyal to all of these brands available out there.

But I do think that you know, superiority in terms of product offering, in terms of service offering, when consumer knows that it could also generate a kind of loyalty that may not necessarily involve a whole lot of emotions, but cognitively they know you're better.

SS: Let's understand too, the foundational drivers of loyalty are - I mean obviously having great products, great customer service, high perceived value, price competitiveness. Those are all kind of table stakes, aren't they? But you've got to get those right in order to win loyalty. But what are the other - you referenced CX as an example - what are the other drivers of loyalty that matter or need to be considered here? [19.40]

YL: The elements that you've already mentioned before. The having a good product, having a value, price-value standpoint, that you are doing a good job with, that you have a good customer service. Those are fundamentals. I would probably say that if you don't have the fundamentals correctly, you probably don't want it to be even thinking about these other things just yet, right? Because ultimately even if you have other things that are attractive to consumer momentarily, if you kind of lose out on these fundamental components, you probably will not be able to keep people's loyalty. But beyond these components, I think there are other components depending on, you know, which layers of loyalty you're talking about.

So I think about two years ago we did a study essentially looking at customers in four different countries in the United States, in India, in UK, and in Australia. So when we look at that, different generations, people kind of offer different type of things. But overall if I were to say like, what kind of factors they are, some is about the experience that they have. So I'm talking about beyond the standard customer experience, right? You have good customer service, that's one thing. But there is also experience component in terms of, do your brands provide a fun experience for me? Does your brand provide a personal experience for me, that where it makes me feel like you understand me as a consumer, right?

And so there's also, does your brand become a connector for me and others around me to create that feeling of social connection, right? A lot of the luxury brands, very community-oriented brands, kind of create that where you feel like you're in the, in a selective community of people who are in this space.

More recently, I think our research on the identity alignment as well, right? Do you feel like this brand is representing who you are? Now these are typically not necessarily every brand will follow that kind of example, but some of the brands have been successful because they really resonate with consumers in terms of who they are.

SS: And values.

YL: Yeah.

SS: Alignment of values. I think Patagonia would be an example of that. Just in terms of the social cause, sustainability, et cetera, draws people in because of that. For sure. So given that, are there generational differences here, “Boomers” versus obviously “Gen X” versus “Gen Z” (‘zee’ in the US). Are there generational differences that you've identified as well that need to be taken into account?

YL: Yeah. So, what we have found in our own research is that generally what we talked about as kind of value drivers, value in the sense of your worldviews, that type of value, tends to be more important to the younger consumers. I think they grew up in an environment where there are so many brands available and more brands are engaging in more differentiating process because the typical product mix, the price and all of that are fairly comparable across multiple brands. And so for these kind of younger consumers, I think that that value component, whether you care about sustainability, where you, whether you care about, you know, the human race in general, you know, would play a bigger role in their behaviours. When you are getting to the older generation, you know, “Boomers” and older, what you're looking at there is that those fundamental core drivers of price, for example, or product and so forth, would play a much more important role.

And so, we're kind of seeing the market to kind of shift from the original, that very traditional driver of loyalty type of picture to a much more diverse, heterogeneous, in terms of why am I loyal to a brand? I could be loyal to a brand because it's really fun. I love my experience, whether that's shopping in the store or whenever I interact with a brand on social media, right? For another consumer it might be, I really care about nature, are you having sustainability initiatives? Are you having ethical business practices? That's what I care about. It's more fragmented, I would say. And brands are needed to consider different spectrums and determine what is the position they should be taking.

SS: And we're going to get into loyalty program design in a second. But certainly that has to be taken into consideration if you're thinking future forward in terms of your program design. I want to come back to that because it's a big question. I do want to end this sort of segment of the conversation around, I want to go back to what brands are trying to achieve here. And a lot of brands are, you know, seeking “brand love”, quote unquote. I'm going to come back to that as a, as a measure later. But what should the goal be here?

If you go back to as far back as 1923, the first person I think define, to define loyalty is Melvin Copeland3. Correct me if I'm wrong there, but his idea was that loyalty is when the product is not substitutable. That is, no one's going to switch products because they are 100% committed to that particular brand. That seems extreme. I know in CPG as an example, loyalty is defined as every other purchase, I believe. How high should brands set the bar? I presume this is also somewhat category dependent. But what's a realistic goal here? If you've got all these gradations of loyalty as you were describing earlier, what does ideal look like? Is it 100% committed to the brand? [25.30]

YL: You know, that's a, that's a good question. I would probably say every brand where they wanted to be is probably what you were talking about, right? Completely committed to the brand. If I were to buy a phone, I go nowhere other than Apple. If I were to, you know, buy a, some kind of grocery item, I only buy from this particular company. And the consumers actually also love you. Not just out of convenience, out of these other things. That is the ideal picture. But is it really achievable for all of the brands out there? I'm not sure that would be the case.

As I mentioned earlier, we all have certain capacities as humans. You know, our emotional connection capacity. There's actually, I think some sociology study that looking at individuals, like, how many friends you would be able to have, like true friends, right? And it's not an unlimited amount. Even though we all crave friendship, crave social connections, but there is capacity for individuals. And so, I think the brands that can reach that level for an individual consumer, not necessarily at the market level, but just to think about like for a consumer, how many brands would I be able to really say I'm truly loyal to in the sense I absolutely love the brand, I would go nowhere else other than I shop at those brands. I would say that's a limited set for most individuals.

But at the market level, what you're potentially looking at then is that you need to audit your own brand, audit your own products and to have an idea of what, what am I good at if it's not authentic for a brand? That's a very traditional type of model to all of a sudden to say we are a social value brand, right? And so, there's a, needs to be a little bit of that or a good amount of that authenticity in there, in what it is you do you have to kind of be honest with yourself, be honest with your market on what is possible for you.

Now if you happen to operate in a product category where there's a lot of consumer taste, for example, needs to be very variety seeking. I don't want to eat chicken every day, right? I wanted to, you know, diversify my taste. When you are in a high variety seeking environment, then perhaps 100% share of wallet is not achievable because nobody wants to necessarily have exactly the same thing every day or I should say 99% of the market. And in that kind of environment it's maybe potentially how frequently people are coming back to you out of that possible set. People are switching between, are you kind of the primary choices for consumers out there? So it's a hard question to just give a, I would say one single straight answer because a really different brands their situations are so different.

SS: Yeah. And that's the idea of establishing both category benchmarks and historical benchmarks and obviously continue to try to improve that. I want to touch on something you were talking about earlier in terms of the different types of loyalty, cognitive, emotional, behavioural. I mean the intersection of that really constitutes ultimate loyalty, I suppose. But most brands today, and go back to the brand love question, aspire to develop that emotional connection. I mean the payoff is obviously in the repeat purchaser. But ultimately they, I imagine every brand wants to be the Apple of their category, just that blind devotion to the brand. Is emotion the critical factor in sustaining loyalty over time? Is loyalty just a fragile thing until that emotional connection is established and firmed up? [29.13]

YL: That's a good question. I will give you an example. My own personal choice, right? So there are two brands that I'm really loyal to. One of them is Lego. Lego, I'm really loyal to them just, just because it's. I'm, I'm sort of a nerd. And you know, Lego kind of creates products that are perfect for nerds, right? You, you kind of putting things together and you, you have so much fun in that process. So for that particular brand, loyalty is a very emotional experience, right? It's purely built, built on emotions. I don't think there's, you know, there are plenty of mimicking products out there, that you know, create things like you can also put together and all of that. But the way I feel about that brand is a very emotional experience.

Versus when you're looking at, I mentioned earlier, like Pentel pens for example. I don't necessarily feel emotion about my writing instrument. It's just something that if I, you know, it's, if it doesn't work, it annoys you. So I guess from that perspective it's somewhat of an emotional experience. But mostly I just know having the knowledge of that this product just works better for me. I, when I use it, it just feels good. So I don't necessarily think that emotion is a single point for consumers why they are loyal to your brand.

And I don't also, don't necessarily think that that emotional component of it has to be there in order for that loyalty to be sustaining. But what is challenging about cognitive loyalty only though, like where I know your product is superior is that you can always, you're always at the risk of somebody else come up with a potentially superior product, right? And then at that point you're potentially going to lose customers. So for me to switch from that brand that I'm cognitive loyal to, it's not necessarily the price. Somebody offer a better price, I'm not going to be attracted to that. However, if somebody were to offer a superior product, and if that's the reason I'm loyal to a brand cognitively, then yes, that's potentially going to attract me away.

SS: But to your point, with so much product parity, it's getting harder and harder these days to differentiate just on those features. Therefore, you've got to find other ground to fight on. And that's the challenge I think here for a lot of companies earning greater loyalty is they're not good at doing all the other things that tether you to customers, except for reward programs. And I want to spend a fair amount of time on this.

It's my read and I've had enough experience in this industry that most companies treat loyalty as a program, not as a strategy, that often these programs exist in isolation of any kind of coherent customer experience strategy. It seems the default position, if you look at almost every program that's been out there since the early 80s is a digital version of trading stamps. Often customers end up loving the program because it's so well orchestrated and executed, but not so much the brand. Do you agree with that, that most brands, most marketers confuse the two, that they really haven't figured this out yet?

YL: Yeah, I will say that it's really interesting that I came from more of a academic world, right? So when we talk about loyalty, most people kind of have an idea that this loyalty has two different layers, right? And when I had conversations actually, you know, at the beginning time for the Loyalty Science Lab with industry, I'm, I was surprised at how many people, the minute I talk about loyalty, immediately people are defining that as loyalty programs. And so they thought I was talking about loyalty programs.

I was very surprised by that in the sense that I thought customer loyalty and brand loyalty would be a given. That you, as brands, you know, that is an important thing that you need to cultivate and that is so much more than loyalty program itself. And at the same time I would say that, you know, loyalty program actually has a lot of value actually, you know, if we call that small “l”, right? To have the ability to actually build a big “L”. If you are going with your loyalty program, you are designing a while you are doing a good job, engaging customer in the process and all of that and really are connecting with your brand's value proposition, doing a good job with that, then you could actually also grow that long term loyalty towards your company or towards the brand.

But it's sometimes sad to see that that potential is not really kind of utilized. A lot of the business decision to say we need a loyalty program is because we're losing customers to another competitor that's offering loyalty program. Or maybe everybody else has a loyalty program, therefore we need to have one too. What is really the value of loyalty program? I would say it's, it's not just fend off competitors. There's a whole lot more. Whether that's growing your existing customers, whether that's gathering more useful information about your customers or you know, kind of align with your brand value, creating a community - so all of those things, not all brands are necessarily there with their loyalty programs.

SS: I think the issue is that a lot of what you're describing is absolutely the case. The me too aspect, obviously the worrying about losing market share, et cetera. But given the ubiquity of loyalty programs these days, I think it's more the latter. If they're doing it, we need to do it as well. But let's go back to the objective here, I guess a little bit, is the purpose of a loyalty program is to presumably create greater loyalty to the brand as opposed to, as I say the program – is the fact that the very existence of the program blinds management to the things that they actually need to do to improve loyalty. In other words, you know, the use of tokenized currency to buy loyalty basically rather than earn it. [35.23]

YL: I would say that I have seen a variety of purposes for having a loyalty program. Not all of those purposes start with loyalty in mind. One example would be when you think about companies that start loyalty program as a substitution for promotions. So for instant discounts, price discounts and promotions. Because when you gave instant price promotions and discounts and so forth, it makes your brand image looks cheap.

Academic research has shown it has negative long term consequences. And so instead of doing that you use an alternative currency, you do a delayed discount in a way, that where consumers have to earn that reward or discount or voucher, whatever it is that you're offering as a reward over time. So it helps from that brand image perspective. Doesn't necessarily increase your loyalty with that objective. It might, but that's not necessarily your main goal.

So I think the that's one and I've also known other programs where the main primary purpose of having a loyalty program, or at least a starting point of having the loyalty program is to be able to get customer data. You know, brands that usually don't have a direct customer relationship with the brands, they may go through retail channel, for example, and they have to rely on retailers to feed them, you know, their information, which may or may not be a complete picture of it. You know, they design a loyalty program to be able to capture that. So ultimately that understanding of the customers, it should lead to higher loyalty because then, you know, you know how consumers are buying and consuming your products and you get a deeper understanding of your consumers. So that can potentially lead to loyalty.

With all of these different industries and different programs that I've seen before, some of them are essentially substitutions for discounts as a promotional tool. That's where they would probably classify, similar to coupons in a way, just works in a different manner. Some of them are really, truly to make the best customers to feel rewarded. And so in that sense, to create that higher level of loyalty and resonance with your customers. And then some of them is, as I mentioned earlier, is the way to kind of grow customers, to grow the market share, you kind of getting people to concentrate their purchases with your company and so forth. So it's multiple purpose and each program depends on what it is that you are trying to achieve.

SS: Yeah, absolutely. And I certainly see that. I interviewed Richard Schenker a while ago and he was one of the architects of the most popular program in Canada, which is the PC Optimum program. When it was initially launched, it was a tremendous success, patterned largely after Tesco, which pioneered the use of obviously data to drive consumer purchase. It's worked very well for the grocery trade. But if you, and I draw this example all the time, if you look at PC Optimum and its success at 17 million members and the relationship that its parent Loblaw has with consumers, there's two different stories that play out. They make brilliant use of the program to drive incremental purchase. No question about that. However, the relationship between Loblaw and the consumer is not a good one. Their NPS is underwater. People hate the brand, but they love the program. And this is the conundrum I think that you see out there is that - you were making the point earlier about, it's a superficial relationship that exists and it blinds the company to the things they actually need to do to improve the relationship over time. And the thing is, this is the other factor I think you haven't mentioned is that, you know, we live in a society today where you know, three companies own everything, right? So it's not like you have a lot of choice either, right?

Especially - in the grocery business in the US is a lot different. You got Trader Joe's, you got Costco, you got Aldi coming in, as well, you got a lot of really top class CX companies playing in the space and a lot of grocers that aren't that kind of don't, don't get the picture - in that same situation, scenario doesn't exist here.

I just want to ask one other question here because it seems everyone has a loyalty program these days, pick a category, there's a loyalty program. Does it make sense for every brand to have a loyalty program or are there certain situations where no, this doesn't make sense. Why would we do that? Why would we give away margin if we don't have to? Let's figure out another way to earn loyalty. Under which circumstances would you not consider a loyalty program? [40.04]

YL: So when you think about what a loyalty program is capable of achieving, right? So we already talk about the data component of it, we talk about a potential market concentration component of it, the brand image perspective component of it. In some situations maybe you are, let's say you are everyday low price, okay? Your main image, brand image, what consumers love about you is everyday low price. When you have that kind of situation, discount is consistent with your image, right? Offering discount. If you also happen to be, let's say you're an online retailer, you already capture customer data when they make purchase with you. So that data component of the benefit is also minimal.

And so when you think about these kind of components, none of them are really going to add much additional value to your business. Then I would say “don't” because I mean in some cases loyalty program management is complex, right? You need to manage the data, you need to manage the rewards and all of these complex mechanism and you have to think about liability and these other things. So is it for everyone? Do you have the resources to be able to manage those things? I don't necessarily think so. So I do think that any company that's trying to start a loyalty program really needs to think about, these are the couple of things loyalty program can do. Does any of these actually benefit our business context? Only if your answer is a clear “yes” – and ideally across multiple benefits, right? That's when the loyalty program would be considered to be a good tool to be looking at.

SS: So it's creating an undeniable business case is what you're suggesting that can be presented to management to say, yes, this investment is going to pay off over time. Incidentally, a tough one today still to make for a lot of companies to prove to boards that that's the case. But let me ask you then on that same subject, um, the option of joining a coalition program versus operating a proprietary one, what's your perspective on that? There are some very healthy, now healthy, coalition programs here in Canada, Air Miles, now owned by BMO. They've just rebranded that to “Blue Rewards” I think. And you've got the “Scene” program here, you've got a number of them that operate very successfully. When does it make sense for a brand to consider a coalition program versus a proprietary one? Does it go back to what you were just saying in terms of the business case here?

YL: If your primary objective is to increase the engagement with your customers and to increase that emotional component of loyalty, really create that engagement, I don't think coalition loyalty program is the way to go. Because coalition loyalty program, where that loyalty is coming from is really, as you mentioned earlier, there's a difference between brand loyalty and loyalty to the program, right? And so coalition loyalty program, the really well run ones are typically the ones that are, you know, you have very strong loyalty towards the program. I would use your business because you're a part of the program. But if you're ultimately - you're really trying to increase your brand, not just to increase sales, but to increase the ultimate loyalty towards your brand, I don't necessarily think that's the best context to be encouraging customer loyalty.

However, as we mentioned earlier, loyalty programs are not cheap to operate. You know, there are various, a lot of different mechanisms that need to be in place for you to really benefit from a loyalty program. And so, from a small business, smaller business perspective, is that worth your investment based on the, what you're trying to get out of the program? If you are trying to do nothing other than substitution for discount, for example, that's your main objective, then it probably doesn't matter as much whether it's your currency or whether it's a coalition loyalty program's currency. And so there is a lot more substitution over there that you can leverage without having a damaging effect on your brand loyalty. So it's really also depending on the business case that you are trying to have your loyalty program for.

SS: I want to move the conversation into measurement, big subject, appreciate that. But NPS is the beacon measure for just about every major enterprise going, has been for 20 years or so. What's your perspective on NPS as a loyalty measure? And does it offer a true picture of how customers feel? And are there other measures that you favour over NPS as a true indicator of that intersection we talked about: behavioural, cognitive and emotional loyalty? [44.47]

YL: Yeah. So I think there has been actually some academic research looking into the NPS measure and looking at the validity of NPS as a generic gauge of customer loyalty. Mixed findings I would say. NPS is so popular in a way in the business world because it's really easy to implement, right? You literally have one single question and that's all you need to know. Does it reflect loyalty? I think it does because when you think about one of the indicators of loyalty is your willingness to share word of mouth about the brand. And that's essentially what NPS is trying to measure. But is that the only indicator of loyalty? Not necessarily. Because that depends on your customer segment. Not all of us is necessarily telling other people about the brands we love. And so, if you happen to be in the marketplace where a lot of your customers are like that, your NPS score might just look lower. But that doesn't necessarily mean that your customers do not have strong loyalty towards you.

I'll give example of a very niche brand. My son is very loyal to Warhammer. I don't know if you know, they have actually been around for like 30, 40 years and I just learned about it because my son started to really getting into it. It's a very core community of people who you know, assemble these characters and play the games and is a very strategic game and all of that. But if you were to ask more generically about the NPS, they're not going to just go talk to a regular friend who has no idea what this thing is right about it. Their NPS score may not necessarily be high.

But more broadly I would say in academic research when we talk about scientific rigor, one of the big thing that we would uh, academic reviewers would often raise problem is when you use a single item to try to measure a complex construct. And when you do that you are typically simplifying things and you are typically missing out on a lot of important things. And so we typically ask for having multi-item measures that represents different facets of that particular construct you are trying to measure. And so, is NPS a indicator of loyalty? Yes, it is. But is it the only indicator of loyalty that represents the true picture of your customer's loyalty? I don't think so.

SS: I think the other abuse of NPS well, there's a number of abuses of NPS as we all know, but one of them certainly is, it tends to be used as a satisfaction metric, not necessarily a loyalty metric. Every service interaction seems to create a “would you recommend” survey after the fact. But I just want to go back to what you were just saying though, because I think, appreciating the fact that NPS is important to consider as part of the dashboard. But is it important to create a composite score that captures the levels or gradations of loyalty we talked about at the very beginning of this conversation, basically the degrees of loyalty so that you can find the people who score you 100% yeah, I don't buy any substitutes. I'm all in. Whether whatever the price is, whatever premium I have to pay, I'm willing to pay that. Is that of value to augment NPS with a composite score of some kind?

YL: Absolutely. There's absolutely value in having a more refined view of loyalty. With NPS score for example, okay, let's say, what are the things that you can actually move the needle with NPS? You're absolutely right. A lot of times the NPS surveys that we receive is after each service encounter. So if I happen to be very happy with the service, it's going to give me a higher NPS score. So you can certainly change your customer service strategy to make sure that people are more satisfied with it. But in the long run, is that the only thing that matters and how much other actionable things you can potentially do with that just one score? I think it's potentially limited, right? Versus if you take a more, whether you say composite view or maybe more multi-faceted view, you might see that this customer, let's say we have three different loyalty components, right? Based on cognitive and so forth. You might have a consumer whose these two scores are higher, but this third one is low, or this one is high, but the other two are low. Then I think you create a much more nuanced view of your customers and there are more things that you can do business-wise, more actionable things you can take in order to change how that consumer relates to your brand.

So I definitely see the value in doing that. I'll give you a quick example of a project that we did with a car dealership. Over there the loyalty is actually quite complicated because it's are you loyal to the car or the car brand or are you loyal to us as a dealership? Are you loyal to us to buy a car or are you loyal to us to get the service that you need for your car? So, so when you have these complex relationships in that situation, they're really just one NPS is not enough. [50.00]

SS: Yeah, totally. I, that's a really good example by the way, because it is multi level. If I'm happy with the product, I'll buy it again, but might not buy from that dealership because I don't feel I was treated fairly or equally. But it also factors into whether I think that car company is a good company or not. So it's a really, really excellent example. The other thing too about taking a more nuanced approach to measurement is it creates segments that can be treated differently. I always think of a bell curve where 60% are sitting in the middle and you want to pull across the people who are most amenable to developing a deeper relationship. And the only way to really understand that is not behaviourally, but it's actually - go back to emotion – it’s how they feel about you really more than anything. And then the other thing I think about is that 50% of decision making in most cases is in fact rooted in emotion, not in a cognitive dissection of what product is actually best. So it all comes back to segment management to some degree here, doesn't it?

YL: Absolutely. The uh, so consumers are heterogeneous. Marketing science research that based on empirical data have time and time again, proved that when you consider consumer heterogeneity, your analysis of the data, your predictions, your model fit and all of those are significantly better. You know, human beings are just different from each other. We have similar underlying motivations, but at the same time, what makes us tick, each individually, is quite different from each other. In order to identify that, I would say one of the big things that companies are need to, you know, move those middle 60% of the customers you were talking about is to ask the question of why, right? Without understanding the why, it's challenging to really know how people are different from each other. You can't observe their behaviours all day. It's not going to give you the full picture. So I think that in depth understanding of the why with the customers will greatly help with that segmentation component.

SS: Well, and it comes back to - if the purpose of increasing loyalty is to keep as many customers buying from you as possible over time - it's sort of the, the leaky bucket. People don't quit overnight, they quietly quit often.

YL: Yeah.

SS: Often they just continue to reduce their buying. And it's harder to observe that over time if you don't have a longitudinal vision. Which takes me to the uh, question you can't possibly do justice to, but I'm going to ask it anyway. So we talk about loyalty strategy and talked earlier versus program strategy. Would companies be better off backing up and saying no, what we need is a life cycle strategy so that we can actually understand the weighting that we're going to put against the ADR in that life cycle strategy and invest accordingly. Isn't that, comes back to the money vault and where money is spent. And today there's way too much money still being spent on acquisition versus retention. But even within the retention bucket there's development of relationships versus the actual pure retention. So is the problem here if you just take it up to its highest level that marketers generally and companies generally don't have a life cycle strategy, don't have an EX strategy that connects to the kinds of portfolio management we've been talking through about this whole conversation.

YL: I think there are companies that have these kind of customer life cycle strategy based on the evolution of how a customer's relationship with a brand goals, right? So there's a classic research piece from the 1990s looking at brand relationships and it goes through a growth stage, eventually goes to a, you know, dissolve the relationship. So those type of situations happen. Happens all the time. I think some companies are pursuing that kind of strategy, but a majority of the company probably are going more towards consumer to consumer differentiation.

So in other words it's more about which bucket do I put you in and I invest accordingly based on that particular bucket rather than taking a more longitudinal view of the same consumer. How are you growing over time? And what is your opportunity or potential for growing over time or maybe decline over time and then customize my strategy according to that. So that's, I would say, compare with like the segmentation of across customer segmentation is much more rare.

SS: Yeah. But my view certainly, a linchpin really to managing the portfolio, because those relationships evolve over time. It's not static, right? It changes for a whole bunch of reasons. We don't have a whole lot of time left. And I want to ask you one other big question you can't possibly do justice to, but I'm going to ask it anyway. What do you see as the evolution of loyalty going forward? And you know, AI makes everything the crystal ball pretty fuzzy these days. It's hard really to fully understand its impact just yet. But knowing what you know today, where do you see loyalty strategy and management? How would you see it evolving over the years ahead? [55.17]

YL: That is a really big question I would say. You know, we are actually trying to start a research initiative that, actually trying to look at, to that but without the findings for it to be able to share with you at this point yet. I would say that I foresee challenges for the loyalty industry going forward because of the AI. And we have a lot of conversations now about, you know, AI creation, content creation, for example, being able to create ads, be able to write copy for you and, and all of those components. I think it's already pretty developed, right? You can give the right kind of prompt in order to get that information.

But a big part of that true loyalty piece lies in that emotional connection. How do you maintain that emotional connection when you, one is potentially your content is created by AI. Two is also as we see more and more of agentic AI that's acting on behalf of consumers. Then it's no longer with you and the consumer, is somebody in the middle as the gatekeeper and then you and the consumer on the two ends over there. How do you create that and being able to go through that gatekeeper and still being able to connect with that consumer?

So I do think that that future of, kind of, AI mediated type of brand customer engagement piece is going to bring a lot of challenges and obviously we don't have very developed practice on how to deal with that kind of issues at this point yet. But I do see it as a significant challenge for the loyalty industry needs to conquer.

SS: Well and some of the research I've seen has suggested that loyalty has actually declined. NPS scores declining, CX score, SAT scores declining or stagnating. All this money gets spent on programs. 15% growth in Canada alone projected over the next year. Worldwide 17%. Companies are still spending on loyalty programs. Will they continue that trend? I mean if those results keep going south, it's hard to make a business case for continuation of all that spending.

YL: I would say it's probably an argument between the CMO and the CFO there, right? The CFOs are always going to be looking at the, these kind of investments and think of it as potentially a cost center. I don't necessarily see loyalty program to decline or at least not in the short term I should say. You know, I can't predict what's going to come 20 or 30 years later. Maybe at that point nobody has a loyalty program anymore. That's entirely possible. But in the short term I think people are used to the idea of this alternative currency and to take it somewhat for granted that I earn these kind of currency with my purchases so I don't necessarily see that component going away.

The bigger question is how do you match that with the newer AI-related investment that you make? For example, can you, because you have a loyalty program, can you utilize the Gen AI capabilities to be able to understand the customers better? You know there's a fairly recent academic research paper actually used a large language model concept to analyze customer journey and to be able to kind of looking at, so using those methods you can potentially build new understanding of customers. If you're able to do that, kind of build synergy across those kind of investments I think it will continue to be very useful, right?

And so again, just like today without you know, even without any AI, any of that, I don't think loyalty program is a turnkey thing, right? It's not something you just install and you forget about it. Or at least not the way you wanted to really get the most out of it. So in the future I think investment is the same thing. Some programmers will just be running like that. They're never going to really be quite at the full potential. It's just used as something people expect versus the really well run programs. I think they're actually going to potentially benefit from having the combination of today's world with the Gen AI capabilities and together to create more synergy out of that.

SS: And certainly looking at it through the prism of program optimization. Absolutely the case. The risk of AI stripping empathy out of organizations and uh, running totally automated. Not sure that's going to improve customer loyalty but we'll see I guess - obviously an object of your further research and, and certainly I encourage everyone to go to your site and take advantage of the research you've already done. “The Future of Loyalty” report, excellent. The questions about loyalty that you've got are really good I have to say. I just -so I think it's a great resource for the industry. So congrats on creating it and hopefully you can continue to grow that resource base for people because it's much needed and, you know it's been needed for a long time. So thank you for doing that.

YL: Thank you.

SS: And thank you for your time today. This was absolutely fascinating. Time flew by. I have a lot of other questions but maybe we'll set up a part two down the road.

YL: Yeah, thank you for having a chance to chat. I always enjoy these type of conversations. It's thought provoking.

That concludes my interview with Dr. Liu-Thompkins. As we learned, the science of marketing research still has a long way to go to definitively answer many of the loyalty questions that continue to stump marketers. Much more work needs to be done to give marketers the confidence to know that their investments in loyalty building will pay off – that it will drive projectable growth, stabilize cash flow, strengthen pricing power, and result in more profitable earnings – all metrics that corporate leaders care about. These days chasing after new customers to fuel market growth is tougher than ever due to the splintering of media and distracted audiences. Which is why driving repeat purchase and maximizing cross-category buying has become a priority for marketers (according to a recent IAB report). It is an easier path to sustainable growth than battling for mindshare. Prove that correlation and nothing Byron Sharp says will matter anymore. The loyalty puzzle will finally have been solved. But before that can happen loyalty management must evolve from a promotional role to a growth catalyst – from earn-and-burn programs to delivering value throughout the customer lifecycle – from an operational silo within marketing to a pivotal part of the total customer experience.

1 - “How evidence-based marketing works—and why Byron Sharp says reach matters more than loyalty tactics”, Marketer’s Brief Podcast, Dec.17, 2025

2 - “Loyalty Management Market Report”, Grand View Research

3 - Melvin T. Copeland served as Professor of Marketing and Director of the Bureau of Business Research in the Graduate School of Business Administration at Harvard University. He was the first to use the term “consumer loyalty” to describe the relationship between a company and its customers.

Stephen Shaw is the Chief Strategy Officer of Kenna, a marketing solutions provider specializing in delivering a more unified customer experience. He is also the host of the Customer First Thinking podcast. Stephen can be r