Welcome back!

No apps configured. Please contact your administrator.
Forgot password?

Don’t have an account? Subscribe now

The Future of AI and Personalization with David C. Edelman

 

 

Welcome to Strategy Skills episode 492, an interview with the co-author of the book Personalized: Customer Strategy in the Age of AI, David C. Edelman. This book is a playbook for delivering true personalization at scale that will help executives learn how to put personalization at the center of their strategy and accelerate growth. BCG’s Mark Abraham and HBS’s David C. Edelman describe five promises of personalization: Empower Me, Know Me, Reach Me, Show Me, Delight Me.

David C. Edelman is a senior lecturer at Harvard Business School, an executive adviser and board member to brands and technology providers, and an adviser to BCG. Previously, David was chief marketing officer at Aetna and has worked with dozens of companies on personalization, AI, and agile marketing at BCG and Digitas. Forbes has repeatedly named him one of the Top 20 Most Influential Voices in Marketing, and Ad Age has named him a Top 20 Chief Marketing and Technology Officer.

Get David’s book here:

Personalized: Customer Strategy in the Age of AI


Here are some free gifts for you:

Overall Approach Used in Well-Managed Strategy Studies

McKinsey & BCG winning resume


Enjoying this episode?

Get access to sample advanced training episodes


Episode Transcript:

Michael 00:46
Hey, David, welcome to the show. It’s great to have you here.

David 01:32
Terrific to be here. Michael, thank you.

Michael 01:34
So it’s going to be a very interesting discussion based on the work you’ve done over many years, but also based on the recent work that you’ve summarized in a book. But before we get into it, I think it’ll be good for our listeners to get some background in terms of how you arrived at this point. What’s your career been like?

David 01:49
Yeah, sure, I’ve actually been working in personalization for over 30 years, over 30 years, over 30 years. Best, believe it or not,

Michael 02:00
as I probably interacted with some of your work at some point as a

David 02:04
young consultant, back in 1989 before the internet, I was at BCG, I had done three projects in a row that all had to do with how companies should start using customer data that they were starting to collect. Yes, I saw it was pure coincidence, three projects, and I said, there’s a pattern here. There’s something broader. I got some investment money from BCG to dig into it, and came up with the concept that we called segment of one marketing. And I published an article on that, and it hit fire. I was speaking conferences. Clients were very interested, and I began working more and more with companies on how to start mingling their strategies for marketing and technology. Then, of course, the Internet came about, and E commerce, and it became something just much more natural. So I was quite involved with companies as they were starting to get onto the Internet, starting to think about targeting and early, early days of personalization. And then I left to join Digitas, which was a mark interactive marketing agency focused very much on CRM personalization, using customer data, creating new kinds of digital experiences. That was right around 1999 and then was at this, the rock face of working on this with companies actually building new experiences, working with American Express, IBM, SAP, and then I left to join McKinsey and help them build a digital marketing strategy practice which was even more advanced than in how to think about organization issues, technology issues, bigger picture things of moving to personalization. And of course, the markets moved content management capabilities became available. Way more data was accessible, and now we’re at the point with AI that enables just so much more and the ability to do personalization at scale. So most recently, I was Chief Marketing Officer at Aetna, helped to bring a lot of personalization capability, and not just the way they marketed, but in the way they help members become more healthy, trying to promote healthier habits to individuals, and then after the CVS merger, I’ve been teaching, I left, and I’ve been teaching at the Harvard Business School and doing my own independent consulting.

Michael 04:51
So you were the person who coined the phrase segment of one marketing that’s where it came from, that’s right, and you’re the guy responsible for all that kind. Crazy mail I get, targeting me with very personalized offers. So I mean, I do appreciate and at the same time I dislike what’s happening here, but I want to switch gears, because there’s a very flawed view about AI. It’s all about generative AI with chatgpt and so on. But there’s a whole other side to AI that people don’t talk about, that you are using in your work around data and so on. Talk us through the ways AI are used in marketing that we would not be aware of

David 05:31
Sure. So in the book, we actually talk about five different promises of personalization, five different things that companies need to offer to customers to make personalization something that is a legitimate brand differentiator, that isn’t manipulative, and each one of those has an AI component behind them. So first off is know me, know something about the customer, and there are AI capabilities now to do a lot of interesting things with data that people are not talking about enough. So for example, AI, of course, needs data in order to create models, targeting models, triggering models. But AI can also create data. So for example, there’s a company called Data orb that can digitize every interaction that happens in a call center, whether it’s for service, whether it’s for sales. And there’s a ton of things you can do with that data. From a personalization perspective, though, you can understand that a customer called in about something, the tone by which they spoke, the way they used language, whether it was resolved how they felt about it, you can understand what was happening before they called in. All that becomes data that you can use to think about what should your follow up be, or if they call again, you’ve got all that data there, and it’s not just that they call, it’s the whole content and interaction of the call. You can also use AI to bring data together so generative AI writes code. There’s a company called narrative.ai that has a capability where you look at one data set and it understands the schema of that data set a second data set understands that schema and writes the code to combine them and to normalize the data so companies that have had data scattered about in marketing, sales, product, use, service, financial, day billing data, it becomes a lot easier now to bring that data together, to feed it into a CDP like snowflake, whereas before, you would have had just dozens, dozens of people trying to bang out so that’s just a couple of things. There’s there’s others in other areas. But let me just stop there for a moment, just to give a flavor of other ways that AI can help.

Michael 08:16
Yeah, when I talk to executives and so on. Obviously the enormous burst out popularity of chat GPT means there’s a lot of focus on that system’s capabilities, but it seems like that’s the tip of the iceberg, and there’s many other things underneath the surface. I was approached by a company recently to invest in them, and they had this amazing technology where they would listen to the earnings calls, and they would watch the video, and the software could tell them if the executive is anxious, if he’s worried, why is he pausing? When did he use this phrase before? In what circumstance? I was quite impressed with it that you could do that. Because no one talks about that side of AI. And it sounds to me like when I speak to executives, one of the challenges they face is, where do they make the investment? Because everyone comes up and says, I’ve got the latest thing in AI, but they don’t know there’s obviously limited funds. So I know we’re switching gears a little bit here, but let’s talk about that, because that’s the question we get asked a lot. What’s the business case? But more importantly, how do we decide where do we go? Yeah, so I

David 09:21
look at the strategy for investing in AI from a top down and a bottom up approach. So let’s start from the bottom up. Most companies are going to find that there are products and services that are very but narrow, functionally driven things that could improve employee training in HR, for example, things that can improve closing the books in finance. And those are going to be much more narrow, very deliberate, functional things that presumably have an ROI that’s mostly measured in automation. Uh, productivity, savings, efficiency, the ability to scale faster. Those can be funded out of local budgets in the functions themselves. And should have, and should be encouraged to have experimentation. Most of those tools you can do trials with you can start using for fairly low cost. And there’s a whole nother conversation we can have about guardrails and things like that. But let the experimentation happen. Let people start using those. And that can happen at a ground level, at pretty low cost. Then there’s the top down, the broader strategic uses of AI to change the way you do business, and that’s where you’ve got to step back and say, Can this actually change the way we compete? Not just simply basic the basic automation and efficiencies should happen at a functional level, at the top level, can we change our value proposition to customers? Can we change the way we use information in order to change their perception of our brand? Can we fundamentally redefine a segment of the business. So those are going to be broader things to think about, and that’s where you got to think about use cases. And what can you empower customer? And that’s another one of the five promises empower me. How can I empower you to do something? So for example, Starbucks is actually one of the most personalized brands on the planet. They have set up a whole flying wheel focused on their app, where there’s a currency of loyalty points, the ability to pay with just a click order in advance. But a key part of this is understanding you as an individual. Where are you geographically right now? Are you at work? Are you maybe on vacation? Is it a weekend? Is it morning? Is it afternoon? What do you typically buy? Do you buy things on special if they’re promoting all of that when you open the app, affects the interface that you see to make it remarkably simple for you to just touch order and get what you want. That’s an enablement, and Howard Schultz has actually talked about personalization being key to their strategy, key to how they compete, and one of the main factors driving their growth. So I’m just giving that’s an example of a strategic investment to change the way you compete.

Michael 12:55
Now, I’ve never used the Starbucks app, although I have been to Starbucks, right, and I do like Starbucks. They are convenient and clean and so on. But my experience with apps has not been great, because I find apps to be poorly designed, to take a long time to open, but the time you get through everything, it’s going to take about five minutes to do things. But I was with my colleague, my trainer, recently, and we had finished a session, and he said, I’m going to go to Starbucks. And I said, Well, why do you want to go now? You have to wait in a long line. And he said, Well, I’m going to order it in the app. And I said, Oh, that’s going to take even longer. But what surprised me is it took him 20 seconds to place his order. Now that was impressive for me, because most apps I’ve seen are either poorly designed or they’re so clunky and large and take up so much bandwidth we’re transferring data that it takes forever to process a transaction. Now let’s stick on Starbucks, because everyone knows Starbucks, right? It would seem, based on what you’re saying, that Starbucks is using AI, right? What are they doing that’s allowing them to impress someone like me that’s not easily impressed.

David 14:03
So there’s several things that are going on, one of which is what you see when you open so Starbucks runs hundreds of tests all the time on a number of different things, on the design of the app in the first place, on the whole buy flow, how that operates, but also on trying different things for different people, showing them different suggestions of what they could buy, maybe adding on food, maybe moving to a sweeter drink later in the day, and they’re constantly, constantly running experiments. The AI helps frame those experiments, sets up the sample sizes, the cells, gathers what happens, uses that to then feed into suggest. Options for how to optimize what to show on the app, that it allows them to do a massive amount of experimentation at the same time that it’s also then trying to optimize what happens when you open the app. So there is this underlying current that I often talk about, of experimentation being very key. It’s key for Starbucks. It’s key for a lot of personalization, because experimentation is another way coming back to a point I raised earlier, of creating data, the more you do tests, the more you understand how different people react in different contexts, different kinds of people. You’re getting more and more data and more granular data that can feed the AI more to enable it to better optimize.

Michael 15:50
Now I’ve seen a lot of clients. The way they move into AI, at least initially, is they tend to put a heavy emphasis on these chat bots and customer experience and customer interactions, but it sounds your way. Starbucks has operated differently from at least most companies, is they’re using AI for not the product, but the experience development process. So the AI is not the product. You’re not chatting with some kind of chatbot and placing your order. They’re using it behind the scenes to have a better experience and product. Is that a good way of thinking about

David 16:23
it? Yeah. And actually, coming back to your point of an app could be slow. A chat bot could actually be a lot slower than the interface that they’re using.

Michael 16:32
Yes, that’s actually a very good point, because the more data you have in, the more it’s got to read, right?

David 16:37
Yeah. And all you have to do with the starters is touch and it may even be voice triggered, but you just say a couple of things and get it done. That’s a lot simpler than trying to enter all kinds of things into a chatbot or even talk through a whole talk track of what you want in the Chatbot. You just touch a couple of things and you’re done.

Michael 17:01
Yeah, one of the things I’ve seen with clients in the medical space is that when they design their chatbots, they assume people are going to speak in the Queen’s English with no typos, no mistakes, and you’ve got people typing in medical information, they’re making a lot of mistakes, and the Chatbot doesn’t understand there’s a mistake being made. How can it know that? So the idea of having a button you hit on the Starbucks app, it doesn’t seem very elegant, but it solves a lot of problems that you would encounter if you really see how people use things, right?

David 17:33
Yeah, and you see that in Uber Spotify. I mean, a number of digitally native companies have relied on that some of them may eventually add some kinds of chat bot, but the simplicity of just touching and using all of the information about you to affect the simplicity of that experience is what differentiates Those apps. It’s what makes them just a basic part of people’s days. So that’s that’s a big part of our whole stance on personalization is using the data to make it easier to provide new value, cheaper, faster, for somebody to work with your brand? Yes.

Michael 18:23
And we spoke about a consumer facing company, a B to C Company, and they tend to be quite innovative in the spaces. Do you have some examples of companies in the B, 2b, space? Yeah, also being innovative with AI?

David 18:35
Yeah, sure. So one that I’m quite familiar with is a company called verisin, v, e r, u, s, e n. I’m on the board of a venture capital fund, glass wing ventures here in Boston, and one of their most successful Holdings is this company, verison, who focuses on what’s called Mr. O so maintenance, repair and operations. It’s all the things on a factory floor that aren’t part of a product, so spare parts, things for the machinery that makes the product, even cleaning supplies, all of that. There’s $900 billion of tied up capital in MRO.

Michael 19:16
900 billion. Yes, that’s a big number.

David 19:20
That’s a huge number tied up because companies find it very hard to bring together all of the data to manage that in an active way, they need to understand, what are they ordering, what are they consuming? What’s the state of their machinery? What are the prices in the market? Is there third party data outside that might indicate supply chain fluctuations in the future? They often have finance wants to keep the cost down. Procurement wants. To manage at an individual level, but then operations wants to keep the factory going, and you have all these different parties, and nobody has an integrated view, especially if you have multiple plants, what verisin does is coming back to the point I raised earlier about using AI to bring information together. They take all these disparate databases, they bring them together, and then they start modeling scenarios of what would happen with different amounts of supply of different products in your MRO. And so it’s tied to the individual company to their consumption, and they can use this as a single pane of glass to manage all of their MRO buying. And so that’s leading companies to centralize that function more, get a lot more efficiency and start to really think more strategically about where they might want to stock up, where they might want to be more real time, and to do it across everything that they do. So that’s personalization. That’s a way of using AI Gen, AI here to actually manage the code to bring those databases together. But we’re not, you know, trying to use chatbots to do things. It’s a lot of the other capabilities. So it’s a really good example in the B to B sense of where you can really differentiate.

Michael 21:33
Okay, we found our topic we’re going to discuss for the next few minutes, right? This is what I call a drip, drip problem, whereby to a supervisor, it looks like it’s just a couple of $100 being tied up in supplies. To another person, looks like maybe it’s $1,000 it’s not worth my time, right? But when you add up all the drips, you get a river that’s worth $900 billion so let’s look at this company, right? Very interesting example, before AI came along, how were they doing this?

David 21:59
So how a customer would do it? A company, yes, it would have procurement, having negotiated price contracts or supply agreements with mostly distributors, companies like Granger, for example. And then they would have regular ordering based on what Granger and them might have thought is their average consumption. But then prices do vary depending on supply and demand. So sometimes you get price shocks. Sometimes when you want to order, availability may not be exactly there, and it’s also being managed on a factory by factory basis, as opposed to looking at the whole system and where there might be opportunities to get volume discounts, for example. So it was very much, it’s less intelligent, and most of the intelligence was just based on your rhythm of consumption and not much else. So it’s

Michael 23:05
based on the memory of the person who was involved in the transaction or interaction. So this company, was it created during AI, or did exist before the old AI craze came about? Well,

David 23:19
AI, it’s been around for a while. Yes, yes, so let’s acknowledge I mean, it was created before Gen AI before, just based on a lot of machine learning algorithms that were brought together. Gen AI simplified the task because it did a lot of this issue of writing code to combine databases, but seeing the problem and applying Gen applying machine learning type models to better understand consumption patterns and look at bringing data together to create new models of scenarios that technically was a capability you could have had earlier than Gen AI and then Gen AI is making it a lot easier to do things like access the data. I’m

Michael 24:04
going to repeat what you said, because it’s important for the audience just to make sure everyone’s following. So this company, which is doing some very important work in the industrial space, I’m guessing, primarily, they used machine learning previously to do this work, generative AI has allowed them to manage the data and databases better, but they’ve never moved into developing a generative chatbot solution interface as such. Is that correct

David 24:32
yet? But that’s likely to come where you’re going to be able to ask queries, for example, of data using a chat bot, where you can say, what if we did this? Or, you know, I found out we’re putting in a new line in the factory. You know, here’s some parameters to consider. What might we do? So it’s, it’s coming, if they don’t already have it, which I don’t know exactly. Yeah. Um, it’s going to be coming. And, you know, a lot of different systems are going to add chatbot capabilities like that, but it’s the intermingling of machine learning and Gen AI capabilities that I think is important here. I think what’s

Michael 25:15
important with the listeners is that when you use the term AI, we use it as a ubiquitous term to me in this big black box, which for most people, is chat, GPT, but AI capabilities exist on a spectrum. That’s right. And as you pointed out correctly, and I think it’s important people understand this, that we’ve had AI for a while. It’s just generative. AI is something that’s recently become popular. That correct? That’s correct. That’s right. Let’s switch gears a little bit. Yeah, right. I mean, we’re going to be moving around. It’s a big canvas. We

David 25:47
work, yeah, oh, let’s, let’s do it. Michael, the work you mentioned

Michael 25:51
many, many years ago, right? When, as a management consultant, a junior one at that associate and so on, I remember working on a procurement project where we are to help a large aluminum smelter holding company find ways to strip out exactly the kind of costs that this company is so well modeling. So this is not a consulting firm, right? No, no, this is a software business, that’s right. And how do they make money? Do they license the technology out? They license out their technology? Do they take a percentage of the savings as well that?

David 26:28
As far as I understand, they don’t right now, but that is something they are working towards.

Michael 26:33
Okay, very good. So the reason I’m mentioning this is because we have a lot of management consultants in our listening group, and I want them to get a sense of how this technology will impact their businesses, because I’ve seen BCG and McKinsey making huge investments, hiring scientists and so on to build these capabilities in house. That Correct? Oh,

David 26:53
absolutely. I’m building models like the verusen model for various applications in different verticals? Yes, they are.

Michael 27:01
So coming back to the field of personalization and so on, right? It seems to me that a lot of companies are focused on productivity, cost cutting and convenience, but there’s not a lot of element of how to surprise me and make me happy the way that Starbucks happened with that 22nd ordering, what a company is doing. What’s the word I’m trying to look for here to excite customers, to

David 27:25
excite customers and drive growth? Um, that’s what I see as a big strategic opportunity. Is by changing your value proposition, you excite people to drive growth. So there was recently an interview with one of the leaders of AI for IKEA, who talked about new artificial intelligence system where you can take pictures of your room and then swap in IKEA products into that those pictures with the right scale for their dimensions and color and color matching, and start to see what your room would look like with IKEA products instead of what you got now, or instead of an empty room, and it matches all the dimensions. And they are talking about people who use that app are three to seven times more likely to buy

Michael 28:26
that’s amazing. That’s

David 28:27
AI for growth. That’s and that’s because it’s enabling true personalization. This is my room. This is my taste of what I want to see in this room. These are the dimensions that I have to work and then you can play around with things like budget and so that’s changing the value proposition to enable personalization. So I that’s the kind of stuff to think about strategically. And, you know, I can imagine, for example, there’s a couple of scenarios we lay out in the book that I think are going to be happening really soon. For example, go to Home Depot and just an in a chat bot, say, provide a picture of a current bathroom and all of its dimensions. Provide some images from Pinterest of styles that you may like. Talk about just enter what you want in your bathroom, what you’re looking for in terms of budget, ask Home Depot for five options along with the budget for what you could do. Look at those and then tell Home Depot pick one and say, okay, when can you install it? And Home Depot would coordinate all the different aspects of putting it into your house.

Michael 29:50
So Home Depot is working on that.

David 29:51
They are working on those. Well, I suppose everyone’s working on

Michael 29:55
it. The question is, where they’ll actually get there, and we’ll get there. First question

David 29:59
of whether the. Get it, but they are working on those capabilities that would lead them there. And there’s a few different things about that that I think are important to note. One is just using AI to interpret things like a Pinterest image and what that means for the style of the actual products. So you need a lot of metadata and to understand the products that you sell to be able to match that with styles. The other part is the data around the ecosystem of actually putting this in place. So there’s an ecosystem in terms of working with your manufacturers to get all that data, then there’s the ecosystem of the contractors to actually install it. And one of the other things we haven’t talked about in AI is the because of AI’s ability to write the code to move data around, which I mentioned before, I think we’re going to see more ecosystems evolve of companies that are sharing information in order to get something done for customers. I can see it in travel, completely coordinating all aspects of a trip. See it in something like home renovation, a number of things where there’s some kind of project with multiple dimensions that need to be done. Companies will be seen as the destination portal, where you go in and tell them what you want, and they can mobilize an ecosystem to make that happen. It’ll affect loyalty programs, and it will start to create these different rival ecosystems. And I do know that there are lead travel companies that are already working on this, Home Depot, others are working on this in different verticals, because of the nature of the Gen AI capability, being able to simplify movement of data. Of course, there are permissions and privacy issues that have to be dealt with absolutely but with permissions, with privacy and security in place, which is not a simple thing, but nonetheless thinking about these ecosystems as strategic differentiators.

Michael 32:14
I was speaking to venkatsman recently, and he was telling me the same thing, whereby, for a long time, it was the tech companies like Google, who are, for lack of a better word, the best at data, but now with AI, someone like FedEx, which understands the movement of parcels, could be the vertical that’s the go to place for the data and the ecosystem around the movement of material around the United States and other companies would lease it or interact with it in some way, because they would have the best vertical. And you’re saying the same thing, right? That is a potential business model.

David 32:47
Oh yes, I’m definitely saying that. And I think a lot of the larger, more established companies saw what happened in the digital yes,

Michael 32:57
they’re aware. Don’t want

David 32:59
to get caught that way again, they recognize that they have assets in terms of data. The problem has been mobilizing that data because it’s all in different places, and it hasn’t been integrated before, but now they’re starting to see that capability becoming much simpler to use. So they have to worry not it’s easier to bring the data together. It totally heightens, though, their need to think about privacy, security permissions. But, you know, the Googles and the big companies, the big tech companies, have to think about that too. Yes. So that is going to be, I think, a rivalry where we’re going to see in verticals, company, brands, taking the lead as a hub. I

Michael 33:47
think it’s going to be relatively easier, but still hard to get the data together. Because, for example, the data that could have been gotten together, like, for example, in a bank, that’s basically an IT system, and they know how to move data. It’s easier for them to aggregate it. But for industrial companies, you’ve got data sitting in front of centers. It’s a physical process to attach sensors and so on to aggregate that data. So it is a process. It’s not as if it’s an easy thing to do, but whoever gets it right gets a pretty enormous advantage in this game,

David 34:17
yeah. And in industrial processes, the instrumenting of the process has been going on for a while, and data’s being captured and put into SAP systems or other systems like that. So they’ve been starting to digitize that. The issue is then matching that data with procurement data and sales and all of that. That’s where it’s a stretch. So, you know, that’s becoming easier now. It’s not ridiculously simple. No, definitely not. There’s still issues about the cleanliness of the data, and you know, some there are. Always glitches in these processes, but it is easier, and I’ve seen it in action. I’ve seen major companies, Nielsen, Pepsi, taking data from a number of different sources and using tools like narrative to bring that together into one repository that they can now launch analytics and operations off of.

Michael 35:20
It would seem a company like SAP should have an advantage here. You obviously may not know what SAP is doing, but what happened to traditional industrial software companies not been more in the news or leading this? I may be mistaken, but they don’t seem to be leading this. I’m

David 35:36
not aware of where SAP is. I do know of a couple of other industrial companies that are scrambling to start bringing in, yes, bringing in new AI capabilities. A lot of them still have to make sure everything moves up into the cloud, but they get it. They get it, and they recognize that a lot of the intelligence that they can put on top of the data is going to be key to their advantages going forward.

Michael 36:09
So again, it’s a process of who’s going to get there that’s right, and get there in a way that makes money. Now, looking at the name of your book and your work and so on, while you were speaking, I’ve been trying to think of how many CEOs that I have worked with and work with that have told me about the importance of personalization, and then as we go through their strategy, they keep personalization as the dominant theme. Most of them talk about it, but then as we start talking about the numbers and capital allocations, it kind of falls away. Do you also see that? Why is it people don’t seem to realize this is something you’ve got to hold in front of you at all times, and you kind of lose it as you make decisions.

David 36:49
Well, there’s two parts to your question, working backwards. Yes, it first is something you need to keep front and center, and the leaders who are doing it, like Starbucks, like other companies we cite in the book, like Marriott, they do keep it front and center. They have senior executives who are in charge of making sure this happens. And there’s a number of different titles, ranging from customer officer, chief experience officer, sometimes it’s chief AI officer. There’s a number of different titles for it, but they have senior executives who are on it. They are prioritizing data as a key asset, and they are thinking about how everybody, every process, everything going on in the company, can contribute to the blood flow of data, how much more data they can get, how they can use that data in their processes. And they’re they’re asking those questions and putting that on the table and expecting to see progress towards improving the depth, accuracy, recency, of data. So that’s there. I think a lot of companies where it falls away is where they think it’s just a marketing thing, exactly. It’s just a service ops thing, and it’s not something that’s center of the table. When I was at Aetna, I was Chief Marketing Officer at Aetna, and was responsible for building the first customer experience engine that we really had a lot of, which would require personalization capabilities, and what was key to getting the success was making sure I had a regular slot at the executive leadership meetings where I would put things on the table. And it took a village. It wasn’t just my team. I needed to work with everybody, including finance, everyone. You know it, service, operations, sales, the actual wines of business, product. Everybody had a role in this. And we had to coordinate. We had to understand how to make trade offs, budget, trade offs, operations, trade offs. There were decisions always on the table. So I have the space to do that. I do work with companies trying to make this transition, and they don’t have the space at the table for it, and they think it’s just something that happens within marketing that has to be managed out of a narrow marketing budget, and that’s going to keep them

Michael 39:35
limited. It sounds like a very big change management.

David 39:38
It is. There’s, there’s a change management in recognizing the opportunity across the board, keeping your eyes on data. There’s also operational changes I mentioned before, experimentation. And there’s a lot of different aspects of experimentation that are important to making progress. One is just. Having comfort that people are experimenting. They’re trying different things. They’re trying AI tools in the legal department in HR, and let people do that experimentation, as long as they’re still delivering the results. But you want to keep that experimentation, then there’s broader mobilization though, to test bigger things and to think about customer experience and changing that, and having cross functional teams working on that and experimenting. So one large telecom company that I can’t mention the name has reorganized to have five different teams, each focus on a different stage of the customer journey, whether it’s marketing and sales, onboarding, product use, billing and service moves, and they have a cross functional team from different parts of the organization, working together, looking at data which they use an AI tool, one that’s actually now owned by Genesis. It used to be called pointillist, that can pull together an image of the customer journey across the business so stitch together from all the databases where David Edelman had touched different parts of the company, time stamp it and see the actual journey. And using that, they can find out where are their problems? Where do people hit dead ends? Where do they opt out and call into customer service and they’re they’re able to put new capabilities in and see whether people are using them. And so there’s this constant experimentation of test and learn, test and learn, stay on top of the data, understand what’s going on, and constantly improve that. And that has led to their customer satisfaction going through the roof. They’ve had share gains, and it’s from focusing on the team and the purpose from a cross functional perspective with that constant experimentation. So that’s another aspect of the change, and I’ve seen that in a lot of different settings, where AI brings together different functions. You need to have a cross you get the benefit of AI by tying data and operations together across different parts of your business, because you see relationships that weren’t there before, and you enable doing things faster without handoffs. So you’ve got to mobilize to do that, and that’s another aspect of change that’s not easy to do. It’s

Michael 42:42
not easy to do, but if you’ve been in business long enough, for example, let’s take an example of building out a wealth management system for a bank, right? You can’t just leave it to the technology team to do that, because they have to work with the different parts of the business that interact with these wealthy customers to understand the needs of the customers, what processes need to be built into the IT system and so on. But it’s often the case whereby too much is asked of the IT department without giving them first the capability to interact. And something you said that is very, very important. You said that at every executive get together, you had a slot. That’s probably the most important thing, because that’s the start of it all, by making sure that you are able to make the case for why every business leader needs to give you time and resources. Because you need their time and resources. You need their people. Need them to do checks and so on. And I think that’s the part that gets missed. It’s usually a situation where you’re given one slot once a year, make your case, present your budget, and everyone forgets about you, because this is David’s problem. Is that our problem? Right? David, thank you so much. This has been an amazing conversation.

David 43:55
Same here. Michael, absolute pleasure, really.

Michael 43:59
It was such a good conversation that you’ve convinced me to have an offside with my team next week, and we’re going to talk about personalization and whether we’re using data effectively. I think we’re using it well, but clearly there are better ways we can do this. Well,

David 44:13
send me an email and we can take the next steps. I will take care. Look forward to it. All right. Thank you. Michael Ciao, ciao,

Michael 44:23
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’re 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.

Want to learn more about how FIRMSconsulting
can help your organization?

Related Articles