Michael 01:29
Hey, Mark, can you hear me?

Mark 01:31
Yes. Now I can.

Michael 01:33
Mark. I’m going to give you a very important compliment right now, you are the first person I’ve ever seen whose corporate photo matches what they really love. I mean, I’ve never seen that before, and I do many interviews.

Mark 01:47
That’s funny, so I guess I look good then I like, you do look good actually, you caught me on a good day.

Michael 01:54
You know, when I read the work you’d be doing at BCG, which is impressive. So, you know, well done.

Mark 02:00
Thank you.

Michael 02:01
I thought you’d have no hair at this point. You must be traveling all over the place.

Mark 02:05
I have a little bit more gray hair than that photo. It works. I think the sun’s reflecting it in the right way.

Michael 02:12
Well, it’s very good to have you on the show. We’re going to talk about the book is there, right? But it’s more the work you are doing that’s right. I’ve never heard of the personalization practice before, and I want to know the back story. I think it’s going to be very interesting for readers, because when I read about it, I thought, well, this is what marketing should have been. Yes, it became this crazy thing about data analysis, segmenting customers into groups of a million and all kinds of funny things which never worked out. So talk me through the backstory. How did you and your colleagues see this opening?

Mark 02:47
Yeah, it’s been quite a journey. So back in, all the way back in 1989 actually, BCG was the first one to put out a perspective, which we titled segment of one marketing. Actually, my co author, David wrote it, and at the time it was just an idea. Of course, we didn’t even have digital and internet, yes. And then I started at BCG, actually, 20 years ago, and I’ve been focused exclusively on customer strategy and marketing. But of course, even back then, we didn’t have the tools to do this. It really started about a decade ago when we helped Starbucks build their personalization engine, and also at the time, a lot of our retail clients were figuring out how to compete with Amazon and other digital natives and really starting to make big investments in a better way of thinking about the next best action for customers, and that’s how our personalization business was born. We realized at that same time that we couldn’t just help our clients with strategy and implementation. We also needed to help them with the AI data science, technology aspects of this. And so we’ve now built out 20% of our 30,000 strong team globally is actually data scientists, AI experts, data engineers, martech experts. So a lot of our team actually helps augment the client’s team, helps them build out their team, upskill their team to actually get this done. Because it’s been so much of talking about personalization, promising this nirvana of personalization, but not actually being able to execute it. And that’s been the journey we’ve been on, first in retail, but now across every single industry, it’s been a fun journey.

Michael 04:47
You said something very important. So I have a risk background, and I remember the early days when we were doing computer modeling, we had to connect something like 12 laptops together to get a Monte Carlo simulation to run. To get one data point right. And in my mind, I always said, what if we were 510, 15 years in the future, and we could do this for clients on a daily basis, and we actually saw that coming through to fruition. So my question to you is, at what point did you realize it was worth making the investment because the technology would catch up with what you were trying to do. It must have been that gap.

Mark 05:25
Yeah, it was very much about those early pilots. I mean, when we saw that actually getting one to one with in our content, for example, on personalized offers and retail, was a huge use case that we deployed, first to places like Starbucks, but then in grocery as an example, and cloud was emerging a decade ago. Yes, and it’s not a coincidence that we started our practice here in Seattle, where I’m based with working with AWS and Azure and now Google as partners as well, we saw things like eight to 10% revenue lift from personalized offers, yes versus mass promotions, and in some of our clients, we’ve now shifted a mindset of we spend all our money on mass promotions to a good chunk of our promotional investments, 20, 30% is actually done through personalized offers, because you can actually target them in a much better way and get two, three times the ROI from customers.

Michael 06:38
So that first paper about the segment of one that was kind of an initial hypothesis, that was an idea

Mark 06:44
back in 1989 I actually wrote a series of papers done a decade ago talking about what personalization actually looks like and what are the components. And that’s all the thinking that we’ve now distilled into the book personalized, where we talk about the five promises of personalization. They really represent the way to do personal playbook for personalization. So let’s give the audience an example of this. Yeah. So look, it all started with the digital natives. So let’s talk about how Spotify does personalization at scale.

Michael 07:24
Yeah, let’s do that. I have a Spotify account that’s a great, great well, they really

Mark 07:29
follow the five promises, right? The first one is about empower me. What are you trying to do as a brand if you’re personalizing? And Spotify is very clear on that they want to serve you the music that you want to hear the artists that you want to engage with. So that’s how they’ve built the entire app experience. But in order to do that, they need to first know you right? So they need to get you into that app. They need to tag all of your clicks and understand which artists you engage with at what time of day and what geolocation the even the music, the meta tag, all the data on, you know, this is the type of music, this is the genre. This is all the metadata about that music. So that’s all that getting to know you piece. But then it’s about reaching you in the right moment to actually give you the right nudge message or content. So that could be you coming to the app and getting served up your day list, personalized playlist. Could be a nudge a reminder on email or a push notification. Hey, it looks like you’re hosting an inner party on a Friday evening. This might be good music to put on, and that’s the intelligence, the AI really at work. But then it’s also about the show me, you know, it’s so much of this is about the content. Spotify couldn’t do what they do if they didn’t have a library of the largest music collection on the client. So they’re continuously adding to that library and making sure they’re able to serve that up. But also in their marketing communications, they’re continuously tailoring content. And then lastly, and this is probably the most important aspect of personalization at scale, they delight me, meaning, when I come back, the experience just gets better every time. That’s what makes personalization magical. That’s what personalization is. Am I getting a better experience the next time than I did last time? So they’re always coming up with cool, new ways to personalize things. For me, the more I engage with it, the better it becomes.

Michael 09:40
Well, my experience with using Spotify is obviously a pleasant experience. I’ve tried Apple Music and so on, but it doesn’t, I would say it doesn’t know me, because one of the things I look forward to with Spotify is the end of the year. They sent me a summary on what I’ve been doing in the year, and they sent me these really amazing things. They once told me that I’m in the top point 1% Of the listeners of some obscure pop artist in Taiwan. That scared me a little bit, to be honest. And then it sent me another note saying that I’ve streamed this pop artist song more than anyone else in the world. Wow, 186 times.

Mark 10:18
I need to check this fog artist out. I guess

Michael 10:21
I have some obscure taste of music. Let me just warn you, but the thing is, you said something very interesting, and I’m going to touch on this right. I’ve very rarely heard people in the digital space tech, even if I look at the leaders of Microsoft and Facebook, talk about delighting customers. Yes, it’s all about efficiency and doing things faster, better, lower cost. But the delight be part seems to be where the rubber hits the road.

Mark 10:45
There’s a huge miss in the AI conversation right now. So much of it is focused on efficiency. Let’s apply Gen AI and we can get this much efficiency out. Actually, the conversation should be all about growth and personalization is the best way to use AI to drive growth. Our data is very clear on this. We’ve looked at personalization leaders, and they’re consistently growing 10 points faster than the personalization laggards. Let’s

Michael 11:16
expand on this, right? We’ve also seen that with AI. We were talking to an AI provider, and they were telling us, what are the metrics? And they said the metric should be the more people used it. And I said, that’s not a good metric. It has to be about growth. And what I found the most interesting is when we fed it our personal information, it was able to see things about us that we had forgotten about. So as a system, we’re going to invest so much money in you, right? You are the system we’re investing in. Should we do it in this way or this way, or is the part we’re missing? And it pulled up an argument we had made two years ago, and it was the correct argument, he said, but you said this two years ago, and it applies in this situation. So for me, that’s a good example of personalization, but I don’t see a lot of that happening at the moment. Why do you think people are missing that?

Mark 12:04
Yeah, it’s to put up some data against what you just said. We benchmarked hundreds of companies every year in what we call this personalization index, which is the first way to truly get a quantitative measure of personalization. And the data is pretty stark, only 15% of companies are doing personalization well. On this 100 point scale, only 15% of companies are scoring a 70 or above. And so there’s a lot more to do on personalization, and I think it comes down to what I described with the five promises is actually really hard to pull off. It’s the most cross functional initiative you could have in any company, and you need to get the analytics team working together with the tech team, working together with the marketing team. I mean, a good example of this was we did personalization for one of the world’s largest airlines, and we got all these teams working together on a more cross channel way of doing personalization for the first time, right? They were often they didn’t even know each other. They didn’t spoke very different languages, and there was a lot of crying, I’ll be honest, the first month of the effort, because people were massively outside of their comfort comfort zone. So to me, what I like to say is you cannot have artificial intelligence without also having organizational intelligence. So much of this is about breaking down silos, new ways of working. Agile is an overused word, you know, we sometimes talk about, you know, the ceremonies and the quarterbacks and the chapters and how to organize agile. To me, the essence of it is being more agile, but by actually getting eight to 10 doers in a room, working together across these disciplines and getting stuff out the door that actually makes that experience more personalized for the customer, and focusing on a couple of things that today are not personalized enough in the customer experience, and they can make better what

Michael 14:20
You said is very important. In my discussions with executives and consultants, Agile has taken on this unintended definition whereby you’re a fast follower. You gotta be agile to follow. And everyone talks about being agile that we have to respond. It’s become a reactive kind of strategy. And something you said here is that the way I look at it, personalization should be the corporate strategy, and then you align everything behind that. Because the way I’ve seen companies do it, now we do a lot of work for banks, and they have a product strategy, and they put everything behind product, which for now most banks are going after wealth management, and they’ve got a lot of effort and wealth management, and then behind. Have the IT team supporting that. Yeah, and it’s not about personalization. It’s about how do we take share? When, to me, you take share by offering a personal experience to begin with.

Mark 15:10
I love that you brought up banks, because actually it’s the number one area we’re working in. Over the last couple of years, I think they finally woken up to there needs to be a better way. And you’re seeing, you know, for example, Erica, with Bank of America and a lot of the leaders, starting to make bold investments here. I think it really comes down to breaking down these product silos that are literally how banks are set up. So, you know, if you take one of the banks we worked with that was very much organized in credit card and loan and savings account and all these products silos. The big aha for them, and the big step forward for them was when we laid out how long it takes to launch a personalized campaign, which was something like 16 weeks, because 11 teams had to do 15 handoffs, and there was a lot of manual work at each step. And again, it was all driven from a product strategy we’ve got to sell the next product by really rethinking that and saying, Okay, what is the customer journey look like? How we’re helping financial wellness, and we’ve we have the customer has certain goals. Let’s start with learning what those goals are, and then align a much faster velocity way of launching these campaigns. So we cut it down to something like three to four days, because we brought together these agile teams, eliminated the handoffs, and really thought through what those next steps in the journey would be for different customer types. Another good example of this in financial services is fidelity. I mean, they really start with what is your investment plan, which could be done with a wealth advisor or online, and then every step in your next journey is aligned to that. So if you don’t need life insurance, they’re not going to try to sell you the life insurance product. But if you just had a baby, they’ll educate you on the 429, account and the benefits of opening one.

Michael 17:25
What you said is quite it boils down the essence. For me, if you want to make it personal, you have to know the goals of your customers. And if fidelity is putting the goals of the customers first, I can think of it nothing that could be more personal than thinking together with your customers what you want to do, in Fidelity’s case, 510, 2030, years, and let’s figure out the ladder to get you there and give you the right product to take you on that journey

Mark 17:51
Exactly. And that’s what’s powerful about personalization. When you take that mindset, you actually realize it’s not just about selling people things in that journey, there may be, you know, a dozen moments in that 20 year journey, when where they will sell me more. But in order to do that, I might may have hundreds of times when I engage with the brand, where they’re going to learn more about me or educate me on on the flip side, and all of that data is gold when you think about what it tells you about the customer you’ve however, this is where you have to have the right technology and process to feed the data back.

Michael 18:30
Yes, you know, it reminds me of work we do with management consulting partners. I always tell them that if you meet a client and they don’t want to buy, keep meeting them, build a relationship, because the sign of a great relationship is not when someone’s always buying from you, it’s when you keep a relationship and they come to you when they need you, and that relationship should last 20 years, but that doesn’t mean they’re always using your services, right?

Mark 18:57
And by the way, there are moments of truth where they are absolutely not buying from you, but whether you show up or not will define the relationship. So for example, when somebody changes jobs and they’re looking for their next job, that’s a really vulnerable moment. Yes, are you do? Are you showing up there, and are you helping them? Think about their career, think about their next steps. It’s very similar with customers. You know, are you showing up back to that airline? Example, when I lose my bag and my I miss my flight? Are you actually proactively reaching out to me and telling me we’ve rebooked you on the flight and don’t worry your back’s coming on the next one, or are you forcing me to go to the service counter and stand in line and be frustrated and anxious?

Michael 19:48
You know, one really good example of personalization is, I went to a car dealership about maybe three years ago. I wanted the sports car. So, really nice sports car, but it’s a really it’s a. Real driver so it doesn’t have a lane change assist. I’m not a real driver. I want a lane change assist. I want somebody to be beeping and making sounds and I’m going to strike another car. So I spoke to the dealership, I think the manager, for maybe 10 minutes, and then three years later, last week, I got a message from them saying, Hey, Michael, we have a sports car that now has a lane change assist. Amazing. Do you want to come and test out which I’m thinking, what is the system they have behind the scenes that keeps all this information and then follows up on my main pain point? You

Mark 20:36
know, we’ve done a lot of work in the auto industry, and it’s really exciting. What’s about to happen there, I will say most likely that you lucked out with a really great dealership in your case. Yes, I know it’s a very fragmented industry where actually the dealerships are disconnected from the brands, and a lot of the data about customers sits in silos across dealerships and then across the brand manufacturers. And so probably, in that case, maybe had they had a rudimentary CRM system, but it was all about the car salesman’s notes. Yes, their follow through. I think what’s happening now, though, with connected car and the ability to collect data at scale, and the dealerships realizing this, even with things like marketing, marketing initiatives, there’s a lot more collaboration between the manufacturers and the dealers and investment in the data and technology to personalize they’re starting with the marketing piece of the puzzle actually, because only 7% of customers, for example, in the US are in the market at any time. And if you can pinpoint, and you’re able to do that now with actually customer level data, which households are in the market which ones are not, you can massively reduce waste. I mean, think of for the average TV 93% of customers seeing that are not in the market, and there’s massive waste there. You’re getting much more precision targeted ads as a result. And I think with then also being able to incorporate all the data coming back from cars themselves, you’re going to see tremendous opportunities for personalization here.

Michael 22:25
Let’s switch gears a bit to organizational inertia. Why these things don’t happen? These are smart people who run these businesses right the smartest people in the world run banks. You cannot dispute the intellect, their capabilities, the ability to be visionary, see things and plan it out and so on. But if you think of a bank, it has my spending details. It knows how much cash I have sitting in my bank account. It knows that if I subscribe to Redfin, open door in the space of three days, I’m likely looking to buy or rent a house. So I’ve worked with banks for a long time. You’ve worked with banks. They’ve had this data. They know the data is there. They’ve also had the technology. People forget a bank is basically a big IT system when it comes to digital and AI, banks are pretty advanced, more than most industries. Yes, so why has there been this organizational inertia to not offer a more personalized service?

Mark 23:17
I think the number one reason comes down to incentives. Incentives, responsibility of leadership. So when you look at where the PnL sits in banks, it’s with the product PNLs, right? The credit card business drive this many new signups and retain this many accounts, and the loan business has to underwrite this much in in loans, etc. So when you have those kinds of incentives, there’s literally no one thinking about the customer who’s actually got a PNL. The CMO might take a customer centric mindset, might do the research, etc, but if they’re not really owning part of the PnL, they’re not able to drive the change that’s needed. That’s why, one of the key things I stress, and it’s actually part of our the way we score brands on the personalization index, is is there a single, accountable leader who in your organization is a fairly senior leader, at least VP or above, and banks probably SVP or above, and someone who’s owning the personalization PNL, so that means an incremental revenue target and the associated investments against them. That doesn’t mean that person controls all the resources that are good to execute personalization, you’re gonna still need to work with the different product teams. You’re gonna need to work with technology and analytics. But if you don’t have a single senior leader driving that, being accountable for that, it’s very hard, in my experience, to make change

Michael 24:58
what you’re saying is true because. Doesn’t matter what the slogan of the bank is, the strategy, the value system, the way you pay people determines what they will do, right? It’s always been that way. And we just want to unpack this for the readers so they understand this, right? You’ve got a product leader, let’s say the head of wealth management, right? They’re obviously under the pressure to deliver for shareholders, the CEO, the board and so on, and they’re pushing their product out there. At the same time, they have a cost center called digital and technology, which is not asking for meetings, resources, time, which they don’t have, and they see it as an obstacle to them hitting their targets, right?

Mark 25:36
That’s right. And what’s worse is there’s not a great way to measure all this stuff. I mean, even take marketing in general. The CMO might come with, here’s what our mmm model shows. But there’s a lot of skepticism from the head of wealth management or even the CFO on, is that really how much incremental revenue marketing is driving, or not? Same thing with personalization. There hasn’t been a great way to measure how good are we really, and how do we rack and stack versus other companies.

Michael 26:08
Again, it comes down to incentives. The head of technology is not paid. Usually, the bulk of his salary and package is driven by his ability to manage his cost center, and maybe a very tiny portion, maybe 10% if he’s lucky, is driven by the increase in revenue sales from a product. That’s right.

Mark 26:28
But here’s the good news, personalization is actually very measurable, because everything you do is going to be tied to an individual customer and what behavior they you drive with them, so you can set up test and control measurement for anything you do in personalization by definition. So personalization leaders do this all the time. In fact, they measure themselves on how many experiments can they run and how quickly that’s ultimately what companies need to compete on that speed of test and learn in

Michael 27:03
your work. Have you seen that companies have always had the data and now they have the analytics to do something about it, or they’re not going out and tagging and creating new data sets and so on?

Mark 27:12
Yeah. So here’s a common pitfall, which is oftentimes companies tell themselves, we don’t have a perfect data Foundation. We invest three years and all this money to fix our data foundation that is always a path to failure. And every time I’ve come back three years later with companies that have said that just had a discussion with a large CPG and the situation, they’ve never made any progress on the impact? Yeah, you nobody, guess what? Nobody has the perfect data foundation, but what you need to do is pick a use case. So back to my empower me concept. How are you trying to help that customer? And then what data do you have today where you can start to make progress, start to launch a minimum viable product against that, and then what are things you need to add to your data set? Tag more in your data set, there’s going to be critical things to do, but now you know exactly what’s missing, instead of trying to boil the ocean. In that regard, it’s

Michael 28:16
funny. It’s actually quite funny, because, as you know, most of your clients went to business, right? MBAs, yes. They know the case method. The first rule of the case method is, you’re going to make decisions with imperfect data. Yes. But when you’re running a business and, you know, all the pressure, you tend to forget that,

Mark 28:33
yeah. And I think there’s a huge, you know, temptation to just outsource it to the data and technology team, because it’s a data and technology issue. You may not have the technical depth to really know what to do, but that’s where, again, if you don’t bring together business, tech and data, you’re not going to make progress.

Michael 28:56
It seems to me, and correct me if I’m wrong, that the companies that I do very well, will create a framework or an infrastructure whereby they can run a lot of experiments very quickly, at the lowest cost and lowest risk, and then keep adjusting and iterating, keep the ones that work and slowly scale it, throw away the ones that fail, and actually accept that 80% is going to fail. And that’s the way we’re designed to do things. It’s those 20% home runs that we’re going to go for,

Mark 29:23
that’s right. And if you talk to the digital natives that pioneered personalization, whether it’s Spotify or Netflix or DoorDash, they run hundreds of experiments all the time. And the personalization leaders that have emerged that weren’t even digital natives, like you take a Sephora or Starbucks, they’re all doing the same as well. So it’s, it’s a lot about speed of learning, speed of experimentation. In fact, in my own teams, it’s interesting because, you know, in management consulting in general, there’s definitely a some. Skill set around building relationships, managing change that’s important, I think, in my own team, what I look for and where I see the most successful profiles are folks that have that core change management skill set, but also have enough cross functional knowledge. Are curious about the technology, about the data, about getting smart on the various pieces so that they can help clients navigate and act as that orchestrator, that bridge, so you’re getting to faster velocity by identifying what is that next thing that’s broken or next thing that’s becoming the bottleneck, and

Michael 30:39
the data never gives you the answer. You’ve got to apply your judgment to say, I’m looking at this. This is what I think we need to do, right?

Mark 30:46
Yeah, and that’s where there’s, I think, a really interesting evolution to in terms of what good personalization looked like a decade ago versus what it is today. I talk about, you know, and when it emerged, this was very much about great data scientists are building cool propensity models, and now I can I have this campaign so I can take this content and assign the right customers to it, so customers to content. Well now, actually, the exciting thing is, there’s unlimited content you can create, and we can go deep on this with lower cost, faster speed, so content isn’t necessarily a bottleneck anymore. And you can take more of that Netflix approach. Really show me a bunch of different content and but make sure you a, you make it relevant, and B, you measure my reaction to it. So it’s this content to customers approach and really break the mold of well, I it took me this long to create this campaign and this one piece of content 16 weeks now I see if it worked for these targeted customers that I I targeted, and then I can learn the next thing. No, actually, you can create hundreds of pieces of content, run hundreds of experiments across all those micro segments, and then optimize based on those results. Suddenly you’ve shrunk that cycle that might have taken you two years to learn into a few weeks.

Michael 32:19
That makes sense. There’s two good examples here you give the Netflix example, which I think is maybe the best example. They look at what you’ve watched, how much of it you’ve watched, and then they recommend shows based on that interest. And then I have an apple new subscription. Yes, now Apple is a good company. They do amazing products, but when I read an article, I’ve got to click on a little button on the top left saying, I like the article, but Apple should know, I open the article, I scroll through the whole thing, I obviously like it if I read the whole thing, but it creates that extra level of friction. Yes, in terms of personalizing it,

Mark 32:55
I think there’s still a massive opportunity in news and media on really delivering great personalization experience. And it’s actually more complicated, because what you want to avoid is going too far on personalization right and creating just go chamber. Here’s the news that you I predicted based on your party affiliation and whatnot you might like. So when I talk to, for example, the BBC or other prominent news organizations, they’re very much thinking about the ethical ramifications of where they use personalization and where they don’t it might be great to do it in the cooking section or the weekend reading section, and tailoring around your hobbies on the main feature stories you might want to be more broad. So there’s a lot to figure out there, and I think you’ll see more more in that space continuing to evolve.

Michael 33:52
But in a sense, they’re doing it already right. Yeah, of course,

Mark 33:55
personalization, it’s just getting the balance right,

Michael 33:58
because a newspaper can only be so big, they only have so much fun. So the editor in chief has to decide already that we are prioritizing this and we deprioritizing this, yes,

Mark 34:10
and that’s where I think brand guidelines a clear strategy, still have a really prominent role to play. You know, the New York Times needs to decide, okay, if they’re really going to stand for original content and editing, what stories do they want to feature overall to everyone? And then where do they apply much more tailored, targeted content for folks as well, to drive engagement? Keep me coming back, they’ve done a great job with their personalized games, for example.

Michael 34:39
So one thing I’ve seen with the rise of AI is, and I don’t know if it’s good or bad, I’m gonna think it’s a little bit bad, is that the focus on generative AI, and there seems to be less of a focus on machine learning.

Mark 34:51
Yes. So Gen AI is definitely been hyped, and it has a very important role in content. Information. It does do what I described earlier, which is massively reduce the time and cost it takes to create content. And it’s really good in certain areas, like if I’m going to translate the content into different languages or switch out the background to be appropriate in different ads, it already does a great job at that some things are still evolving, like video or showing things like hands. I was talking to one of the beauty retailers. We still have a way to go, but I think people under appreciate how critical machine learning and AI is. In addition to Gen AI, you can create all the great content want. We have a massive risk right now of a content explosion where we already think we get way too much, and especially in this election, but we’re going to get even more, and there’s a massive risk of that, because companies and brands are not pairing the AI generated Gen AI generated content with a smart machine learning approach to delivering it. Already, you have more content than you have custom customers in many brands, in many situations, you need to select only those from them that are really going to resonate to folks. So in another airline example that we worked with, we actually took all the content that they were generating across 50 different lines of business, because they had loyalty businesses, travel insurance, health insurance, etc, they were selling as well, in addition to the airline and we used AI to select what is that next best conversation we wanted to customer. As a result, they decreased the frequency of communication from with some customers it was 10 times a week to just a couple of times a week, and increased it with millions of customers, they were not touching at all because they didn’t show up in their propensity models, as high propensity to like really actually select the next best conversation, and the conversion went up by 3x customers felt better because they decreased the level of irrelevant communications, but they actually increased the conversion. And it wasn’t just in the airline business, it was every single product. They saw gains in that conversion. So and what was cool about that, AI, is it scored every customer every hour, across 1000s of pieces of content, what was the best thing to show and in which channel? That’s the other thing. Where I think the puck is headed is, you know, before you had email personalization and web personalization and in person, channel personalization, acting separately, we did this across the channel so that we knew, Okay, this person we should show this targeted ad to, we should send this email to, or we should show this on the homepage, on the website. I think that’s the next frontier. Is cross channel. Ai driven next best conversations. It

Michael 38:15
reminds me of a project we did once for a credit card company. They would have their call center people be ranked and scored based on the number of calls they completed in an hour, and then, if they were very efficient, they’d get an extra 15 minutes. And but when you ran the numbers through the system, the system said it would actually make more sense for you to work with your customers to solve the problem, because they’re likely to continue spending, yes. And then we ran a pilot, and the system was not completely right, because only customers who spend at a certain level continue spending, but for those customers take the time to solve the problem. For

Mark 38:58
them, yes, you can now solve that equation at the customer level, and not just understand should you be spending more time or not, but also know exactly what conversation to have, exactly like with AI test out different scripts and even the tone of the script and what product you should be recommending and with what action call to action? All of that can be fed in to the AI, and you can arm your sales force with it. I love some work we did in department stores where, you know the sales associates had these, and this was a luxury department store, right? So sales associates are spending lots of time with the best customers, and they were even texting back and forth with their best customers, but they could only touch the top 1% of customers, and with AI, we armed them with much richer data about their customers, not just the 1% But the top 25% of customers, so we knew if this shopper looked online for shoes and but she didn’t find what she liked, the sales associate that knew her could invite her into the store and tell her about the shoe event that was coming up, and also you know, pre built the templates for the store associates. So maybe in the past, they were only able to send out, you know, a few dozen texts a day. Now they were able to send out hundreds of communications between emails and texts and other outreach they made to their customers and touch 20% of the customers instead of 1%

Michael 40:40
in having this conversation with you, I remember banking several decades ago. It was dominated by these MBAs and guys who played lacrosse, and then over the last 40 years, we’ve seen the arrival of, I don’t like to use the word quants, because it’s maybe a bit demeaning, but people that have strong technical skills, yes, rising to the top of banking and banking is completely different from the way it was 40 years ago, even 50 years ago, the pools of profit are not driven by the Goldman Sachs of the world anymore. Yeah, complete sea change. And in fact, Goldman Sachs is not run the way it was run before. But I’m saying the same thing happening in your firm. I mean, you’ve got 1000 agile marketers, data scientists, engineers, marketing tech experts. This is a different consulting firm.

Mark 41:28
Yes, absolutely. I mean, today we are 10x the size I was when we were when I joined. But you know, back then, we didn’t do any AI work, and now a quarter of our work is is linked to to AI, and a lot of it is about building platforms, building reusable systems. In fact, we built a whole reusable code base around personalization that we call fabric, which is essentially in addition to your martech tools like a Salesforce, Adobe and the others, there’s a lot of automation you actually need to put in. And we found our clients having to build that and rebuild that from scratch every time. So we actually now sell IP and assets as part of our work. We are a very different company. You couldn’t have imagined doing that at BCG 20 years ago when I joined.

Michael 42:28
So the journey that you’re taking customers through the journey your firm has actually been on, yes,

Mark 42:33
exactly. It’s been a really exciting journey. Every three years, I feel like I’m having to reinvent myself because the market is changing so fast. And the cool thing about being a BCG is interacting with all these leading brands around the world and getting to experiment with them and figuring out where the puck is headed next together.

Michael 42:58
Hockey of reference,

Mark 43:01
yes, I did play a little hockey back in my high school days.

Michael 43:05
Are you Canadian? Many? Chef? No, no,

Mark 43:07
this was in Europe, actually.

Michael 43:09
Oh, in Europe. Well, that’s good, but a pretty good leagues up there, the Finnish Estonians and so on.

Mark 43:14
Yes, that’s right, I’m Hungarian. Hungarian.

Michael 43:17
Okay, well, that’s interesting. So we’ve got listeners from all over the world, a lot of management consultants, ex management consultants, people in banking and so on. What can they do on Monday morning to understand their maturity within the organization at being as personalized as they could be? Because it sounds like there’s a lot to do here. What can we get them started on thinking about on this journey?

Mark 43:39
Yeah. So I would say a couple of things. One is you need to accelerate your journey and you need to get started now. There is no shortage of ideas in any company I walk into on what to do with personalization, but it is a question of prioritization and just getting going. So pull your team and the cross functional team into a room and just whiteboard the ideas and use cases. And very quickly you’ll identify some where you could accelerate the journey. And even with the current technology, do better get started on that next week charge the teams with concrete goals on how they could get better on that, I think the second, second piece is measurement. Though, in order to know how you’re doing, you’ve got to be able to hold the team accountable to progress. And that could be as simple as setting up the clear measurement dashboard, and that test and learn approach, codifying you know how, how those customers are improving and how you’re seeing growth. Putting that measurement in place and trying to make it faster, more automated is a key, key aspect. And then lastly, score your. Self honestly on where you are on the personalization index. We have a really easy way to do that online. You can read the book and ask yourself the questions along the five promises, but I think it’s pretty eye opening when you look at all the dimensions and where companies are, because even the best of them are have opportunities for improvement. I think

Michael 45:22
the most important thing to remember is that no matter how bad a company may think it is, it has pockets of excellence. Yes, and I think what AI does is it allows you to take those practices and spread them out. And I think that’s very just need to keep that in mind. And there’s no company that’s terrible at everything, there’s going to be pockets of excellence. It’s always a good idea to start there.

Mark 45:42
That’s right. I mean, at the minimum, you have customer data and you have transaction data, and you can take that and do a lot with it. You don’t have it to have it be perfectly tagged, and you don’t have to have the perfect Technology Foundation. A lot of this, especially the testing part of it, can be manual, and then once you see what actually works, what drives growth, then you can go to the CFO and argue for the investments. Well, I

Michael 46:11
was talking to David, your colleague, and also speaker Bill Madison, who I think you may know, and they both were telling me that the easiest way to start this is to just have a conversation with your leadership team. Just set aside some time and say, Okay, what are we doing around personalization? That’s right.

Mark 46:28
And I mean, in the book, we’re very clear about this is a CEO level, yes, agenda item personalization cannot be a CMO or marketing function initiative. It needs to be across the company, and every member of the leadership team has a role to play, even the COO if you think about operationally, I mean what you just mentioned around customer service and all the ways that you need to deliver that personalized experience. If you’re a retailer, in the store or in your operations, more broadly, your you need to change the mindset, change the ways of working in every single team. And so you need to argue for, what is the ambition? What is that North Star for personalization, rally the leadership team around it, but then also get clear on what is the role of each leader at the table in making it happen.

Michael 47:21
I think, you know, there’s something you’ve alluded to, but I want to boil it out yet. It’s important. Personalization is not just about knowing your customer. It’s about responding to that knowledge you have of them in a way that’s beneficial and delights them. Is that a good way of thinking about it? Yeah,

Mark 47:37
I think that’s a great point that you called out on personalization is an overused buzzword, and I think it’s really important to talk about what true personalization is. I define it as taking what you learned in an interaction with a customer and making their next interaction better, faster and more convenient, cheaper and like. So there’s two components there, right? You’ve got to bring back that beta, whatever you learned about them, and then be fast enough to turn it around, turn the insights around, and actually make a difference. I mean, their next interaction might be in a few hours after the one or in the next day. It could be several months or years after but how do you build the infrastructure, the team, the process, to do that at scale for millions of customers across billions of interactions? That’s what true personalization is. Yes, and

Michael 48:32
again, AI makes it possible that you can see everything that’s happening, make the connections and offer something that is better than simply relying on the institutional memory of one customer service agent.

Mark 48:45
Yes, and it’s AI, because you need to run the models to update based on that data coming in and in an automated fashion. But it’s also that OI organizational intelligence to re engineer some of the processes. For example, if currently the way you run a campaign, email marketing campaign, let’s take as an example is this team pulls the customer list, then pass it to this team to create the content, then this team approves the content, then this team passes it to the vendor, and so on. You’ve lost the customer, right? Because they’ve already interacted with you multiple times by the time you turn this around. So how do you shrink that process and impair the organizational intelligence? And that true agile with AI,

Michael 49:32
we’ve spoken a lot about some great B to C examples, other companies in the B to B industrial space, or even governments that are doing incredible things. Yeah, yes.

Mark 49:41
So actually, lots of examples there. So let’s take three of them. One is in the distribution space. So take food service distribution. You know, this company has hundreds of 1000s of small businesses. They serve. Uh, small restaurants and food service accounts, and they’re actually using AI to understand what are those customers ordering and suggesting for their sales force, and in their email and other communication of the customers suggesting what is the next thing that they should be ordering. So they’re thinking through things like smart wreath punishment or items that maybe that restaurant isn’t ordering from them, but should be, because of the broad categories expanding cross selling, and they saw 10% revenue lift when they rolled this out across their accounts because they’re using that data to be more targeted. I mean, Microsoft is another one. Of course, they’re a pioneer in this. Given their AI investments, they set up something called the global demand center, and that team actually scores all of the interactions that they have with key accounts, many different stakeholders and lots of different companies, everything from the salesperson had an interaction to they went online and clicked on this on the website, or reacted to this email, or went to this live event. They take all that data and score, what is that next best interaction the Salesforce should have with that given stakeholder in that given account. So you’re seeing, you know, there’s a lot of talk about how Microsoft leverages AI more broadly. I think what’s really interesting is how they’re applying it with their sales force as well. And then I think government is another frontier to leverage in this. Certainly you saw, for example, Australia made lots of investments in digitizing their public services and making it personalized. And when you go online and try to access certain services, they especially did that during covid, when they were some of the leaders in terms of pioneering digital tools to alert people to covid in the area or contact they’d had, et cetera. But now they’re also applying it in broader public services around tax or opening accounts and the like. I think there’s a lot more to do with governments around the world to applying those kinds of approaches.

Michael 52:26
When you give the first example about the food, catering, ordering business and the analysis that was done, I was thinking to myself that not even a decade ago, that would have been the work of a business analyst, maybe an associate, one stream running over two to maybe three weeks that can now be done in, I mean, few minutes, most That’s right.

Mark 52:49
And, and that’s it’s not just done once, is the power here, but for every account, every hour. And pick up those signals and do it automatically, and you can cost engineer. Now, when we started this a decade ago, we had to be really thoughtful on gosh, the cloud costs get astronomical if you really try to do that. Now, you can architect these models, and cloud costs have come down to the point where it’s feasible?

Michael 53:22
Well, we’ve been doing that in risk portfolio management, portfolio theory and so on. That’s our area that we’d like to explore. And when we fed it all this data, it actually picked out the nuance that we had missed in the way we construct portfolios. And even asked it, are you sure this is edited? Yes, I am right. These are the reasons, and that’s something that would have taken us at least eight weeks of work to do. Yes,

Mark 53:47
I think it’s astounding what kind of insights you can draw on from personalization, not just for marketing, but even new product trends you might be able to pick up on. For example, fashion clients color might be returning as a key trend, and these kinds of colors done what the models are picking up or in beverage customized beverages in the restaurant or beverage retail space are a trend on Tiktok, and how that’s playing out based on individual customers. So the implications for, again, the whole enterprise are profound. Well,

Michael 54:30
it’s a good example. You brought up Tiktok, because Tiktok is almost the entire valuation. Is that system of personalizing videos, because I’ve seen the debate between the US government and Tiktok, and Tiktok does not want to give away the algorithm that allows it to personalize videos. That’s the value of Tiktok.

Mark 54:49
Yeah, it’ll be really interesting. Whoever buys Tiktok will need to actually, most likely, rebuild an algorithm, and they’ll need to do it very quickly. They’ll have a few months. Months based on when that deadline is, and you can have

Michael 55:02
a few angry people or not, yeah, personalized videos. That’s

Mark 55:06
right. I think there’s going to be tremendous pressure to you know that is the core of the experience. I think you’ll have a few a little bit of a grace period, and so you can break it down into, what is that minimum viable product look like. How do you augment it and really transparently communicate it to customers? But you’ve got to hit it out of the park pretty well in the beginning, and that’s the core of the platform, but

Michael 55:35
it’s never discussed in the media, because if you think about it, Tiktok has vastly more content than Netflix, yeah, but if Netflix was not recommending shows to me, I would probably cancel the service, because it’s too much. I don’t have the time to sift through that entire library, click into my remote keywords to search for things and so on. So it sounds like a small thing, but it’s actually, it’s the crux that’s going to unlock the value.

Mark 56:00
Yeah, you have a an amazing you’re buying an amazing base of engagement, but that will quickly shrink to zero if you don’t have the right personalization. Yeah, it’s a

Michael 56:13
little bit like buying Ferrari, but not getting the rights to the engine right technology,

Mark 56:18
exactly. I call it the personalization engine for a reason. Mark,

Michael 56:22
such a pleasure. Likewise, I really enjoyed it. In fact, after speaking to you and David, I thought to myself, we’ve got to have our own internal meeting next week, and we got to have this discussion. Because, you know, as we spoke about, it’s not about having a personalization strategy. That is your strategy, your strategy, your corporate strategy. How do you offer the most personalized service to your customers in a way that makes money allows you to grow? That’s

Mark 56:46
right, every brand is competing on personalization today, whether they like it or not, so got to figure this out.

Michael 56:53
Thank you so much. Mark. I’m sure we’ll have you back on once you’ve done some more exciting work. Thank you. Looking forward to talking take care. Ciao, as we wrap up, today’s podcast is sponsored by strategy training.com if you want to strengthen your strategy skills, you can get the overall approach using well managed strategy studies as a free download. Go to firmsconsulting.com forward slash overall approach. And if you are looking to advance your career and need to update your resume, you can get a McKinsey and BCG winning resume template example as a free [email protected] forward slash resume PDF.