Contextual Engagement Requires Better Use of Data and AI

Jim Marous:
Hello, and welcome to another Banking Transformed Solutions podcast. I'm your host, Jim Marous, owner and publisher of the Digital Banking Report and co-publisher of The Financial Brand. The importance of using data, analytics and advanced marketing technology has never been greater. Consumers expect their financial institutions to know them, look out for them and reward them. Unfortunately, few organizations are adept at using data for contextual engagement. Now more than ever banks and credit unions must leverage the insights they have to create better experiences and higher engagement.

Jim Marous:
We are very fortunate to have Sarah Welch, managing director and head of marketing at Curinos on the Banking Transformed podcast. We'll be discussing how the advances in marketing technology coupled with data, analytics and applied insights is transforming financial marketing. Marketing has taken center stage at the majority of financial institutions, being asked to build communication strategies that build customer engagement and positively impact sales. More importantly, the focus has shifted from a product focus to a customer focus with a power of data, insights and technology driving improved results. As mentioned, we have Sarah Welch, managing director and head of marketing solutions from Curinos on the show today. Before we begin, Sarah, can you share with our audience a little bit about yourself as well as provide an introduction to Curinos?

Sarah Welch:
Absolutely. Thank you so much for having me. So, a little bit about me. I am managing director focused on marketing solutions. And what does that mean? It means that I sort of sit at the intersection of data and technology and insights really to help marketers make that transition or help their banks make that transition from the product and distribution led world to the customer centric, customer driven world and Curinos, as a company, is similarly focused on customer data and solutions and insights that help banks and financial institutions better navigate the shifting sands of today and anticipate where we need to be going for tomorrow.

Jim Marous:
So you're really an advisory firm that helps marketers take what they have at their disposal and really help them make the best use of it then?

Sarah Welch:
We have an advisory business. Yes, absolutely. And we also have data and technology solutions that can help marketers actually implement some of those ideas and activate more productively on those insights.

Jim Marous:
That's great. So let's start from the beginning in a way, what is happening today in the world of bank marketing? What is keeping the chief marketing officer up at night? Probably everything.

Sarah Welch:
What's not happening?

Jim Marous:
Yeah.

Sarah Welch:
I mean, I think you said it pretty well a second ago. It's really an industry that's shape-shifting and it's not too often that you get to play in a sandbox where everything is getting reinvented and in retail banking, everything is getting reinvented. What is money? What is a bank? What are we doing? And marketers in that shape-shifting environment have a really, really important role to play in not only being the voice of the customer and helping bring that into the strategic conversations about where the bank will go and move to, but then also helping the institution move more quickly in that direction of customer centricity. So they, as you say, are playing a really central role in the transformation of retail banking.

Sarah Welch:
In terms of what's keeping them up at night? Probably a lot of things. If we focus on kind of one big area, there's the distinctiveness challenge. So if regionality, the physical location of your bank is no longer a primary determinant of where I should bank, how do you create distinctiveness and define distinctiveness for your bank in a way that stands out? And I know consumers definitely tend to see banks as largely the same. So that's definitely one area that's keeping banks up at night or bank marketers up at night.

Sarah Welch:
Acquisition efficiency. As marketing becomes more and more important in driving new customers into the franchise, in keeping those relationships healthy and deepening those relationships, that is a really important question. How do I use the limited dollars that I have? And nobody feels like they have enough dollars to move the needle. How do I use those dollars most efficiently and almost think like a fund manager in terms of allocating my dollars most efficiently across the markets we play in and across the product set that we have and across the channels.

Sarah Welch:
And then the last is personalization and really that's about relationship depth. How do we productively drive relationships at scale in digital channels as we shift to a more digital first world?

Jim Marous:
So it's interesting. I had a great guest, we had Raja Rajamannar, the CMO of MasterCard. He wrote the book called Quantum Marketing. It's a great book I recommend to everybody, but it was interesting because one of his major takeaways is that the marketing world has changed so much in the last year, but even more so in the last five years, to such a degree that it's almost hard to stay on top of what's going on in marketing. So how does your organization help marketers stay on top of what's happening and actually supplement and compliment their organization's capabilities to do what you just mentioned?

Sarah Welch:
Yeah, absolutely. So a number of things, and I think that there's one central theme and I don't know if you've heard the expression, "From mad men to math men," as sort of being a nice little wrapper of the changing role of marketing. The role is increasingly analytical, but it's sort of one foot in analytics, one foot in still concepts and human centered communication. So we help our bank marketing partners with insights, certainly a lot of research, a lot data to help our clients better understand what do customers need, how are those needs changing and where might banks better play and engage with those consumers? We have marketing analytics that help our clients better understand how to maximize the limited marketing dollars that they do have. And finally, we have a platform that's an AI sort of optimization platform for relationship driving messaging. So for known customers, how do you dynamically interact with customers, with marketing interactions that are more likely to deepen relationships and drive ultimately CLB?

Jim Marous:
It seems in your organization for years has really done a really good job of providing, as you said, the ability to say, "Here's what's happening in the marketplace, here's how to address it," and then give them the tools to help them address it. What's interesting though, and the research that we've done at the Digital Banking Report shows that while we talk a lot about data analytics and AI, the reality is the only areas that financial institutions almost uniformly have done really well in this space is in the risk and fraud, being able to avoid the problems. And that's where AI has really gotten its foot wet in the financial services. But with financial institutions, with regard to using data analytics, applied analysis and AI to move the customer relationship forward, the financial institutions have done a much worse job on that. Do you believe that the organizations that you work with are challenged by this and how are they getting over that hurdle of using data analytics for better customer relationships, as opposed to simply risk avoidance?

Sarah Welch:
You've put your finger on it. It absolutely is a challenge. And I would say it's a challenge across the board for all types of customers that we're interacting with, whether they're neobanks or the biggest institutions out there. And they're challenged in a couple of ways, they're challenged in getting customer data organized and in a place where there's sort of a 360 degree view of the customer that can be leveraged by different marketing machines to drive better marketing interactions, so that's one area where they're particularly challenged. And another area is just agility. A lot of the marketing technology is built on what we call a linear workflow. You need to know in advance, what is the segment that I'm going to address here. For each segment that I'm going to address, maybe I have some hypotheses about A or B or C messages that I want to test and find the one winner that I will then roll out at scale.

Sarah Welch:
That linear workflow really creates a log jam. We know lots of marketers, talk to lots of marketers every day, who have hundreds of hypotheses that they want to be able to go test, but the actual work of being able to go get the data, set up the tests, execute the tests, get the data from the tests to learn what works, what didn't work so that you can adjust the sort of machinery is, that's really where there are big log jams and frustration about how quickly an organization, a marketing organization can move.

Sarah Welch:
And the last is orchestration. If you think about something like next best action, a lot of banks have done a pretty good job of developing complex next best action machines, and that's fantastic. That means that Jim, the next time you show up or the next time you do something, the bank says, "Hey, Jim, the right next thing to put in front of Jim is X, Y, Z product." That's good. However, any good marketer and really frankly, any human being knows that, that's only kind of part of the process. That sort of just product pushing all the time, which is what those next best action engines enable is only part of the equation. There does need to be value adding content and value enhancing content. And what is the relationship between more relationship oriented messaging and product oriented messaging? So the orchestration is complex.

Sarah Welch:
The other piece of orchestration that's really complex is, if I have a portfolio of 15, 20 products, how do I know that I should be sending you this product versus that product? When, as a marketer, I am often beholden to the lines of business and I need to generate X number of widget sales for each product. And so what happens in that environment is I just send more mail or I send more marketing interactions and that may get the widget sales that I want, but it ultimately might be deteriorating the quality of the relationship that I have with the customer and deteriorating their likelihood to respond or engage with things that I send on a go forward basis.

Jim Marous:
So, it's interesting, people say that you have eight seconds to capture a customer's attention and along the line of what you just discussed, but it's not just what you say, but how you say it. And especially as we come out of the pandemic, the role of empathy, the role of actually doing something on behalf of the consumers as opposed to doing on behalf of the product manager that as you mentioned, drives sometimes where our financial assets go towards what we try to push, how do organizations get around that? Number one, I would assume you agree with that concept, that it's how you talk to, but how do organizations get beyond simply the product push into more of an empathetic dialogue that builds engagement and may not initially build sales?

Sarah Welch:
Yeah, absolutely. And if you think about empathy, which is fundamentally a human trait, what does that look like in practice in banking? So I'll take a consumer kind of walking in to a retail bank, maybe to open an account. They may sit aside from a banker and that banker is really constantly throwing out a series of stimuli and looking for how the customer responds. So might throw out a bad dad joke, see if you like humor. Might ask you some questions. And with each bit of stimuli that they're putting out, they're watching for your response and adjusting course and taking all of those inputs in order to kind of figure out, okay, what ultimately is the right conversation to have with this customer and solution to present to this consumer. And how do I present that? Do I present that in a, "Oh my gosh, the house is on fire. You need this right now, based on everything you've just told me," or with a wink and a smile, or in a very sort of 1, 2, 3, this is how you do it, practical way. So that's how humans do it.

Sarah Welch:
If you think about what marketer's challenge is, how do you do that at scale in digital channels? What you need to be doing are a lot of test and learn. Test and learn and iterate. And there are so many variables that you can be playing with. So if you think about tone, you could have direct tones, fear of missing out tones, witty tones. You can have different kinds of images. You can have different lengths of body copy. All of these are variables that you can and should be playing with in order to find the sort of how wrappers that are most likely to connect at an individual level. And what we have seen is a massive response difference between campaigns that do have that ability to adapt to and personalize on those variable dimensions. We see anywhere from 60 to now 125% increases in responsiveness from customers who are getting much more tailored and personalized wrappers.

Jim Marous:
Let's get into the weeds a little bit. I mean, most financal institutions, a couple things play into this. Number one, I've been lucky enough to meet with bankers across the country over the last couple months, face-to-face. And one ongoing theme from everybody is they are overwhelmed by today. So they really can't even think about what they should be doing tomorrow and moving forward beyond keeping up because the keeping up is such a much faster pace than it was before is so difficult. On top of that, almost every financial institution says that their data is messy, it's multiple silos, it's all over the place. Do you help organizations, number one, make unmanageable data manageable and be able to apply it. And secondly, do you help organizations keep that pace of chain, that pace of innovation, that pace of marketing dynamics moving forward quickly when really they are going to have a very hard, difficult time doing it themselves?

Sarah Welch:
Yes, is the short answer. In thinking about customer data, and yes, that's the lifeblood. We like to call it digital exhaust, and if you think about the best customer experiences out there, which all banking customers are experiencing, at Amazon, at Airbnb and Netflix, at Peloton, you name the company, sort of expectations are being set on how companies will be able to take their data and get to know the customer and respond in a personally appropriate way. It is really hard for banks who have legacy systems and cores and data all over the place and an architecture, historical architecture that was product oriented and so siloed data. Yes, getting that data out and organized is a really important piece of the equation for marketers. It is something that actually is a byproduct of one of the solutions that we have, that it also enables the rapid test and learn and personalization of execution at the same time.

Sarah Welch:
There isn't necessarily a separate 18 month march to first get all of your data perfectly ordered in a centralized place. There is a standardized set of data, and we have found sort of most useful is use case focused. So you don't tackle all of the use cases at once, but you start with one use case like onboarding as an example, or utilization or cross sell. And for each use case there's sort of a standard set of data that needs to be organized. You tackle that and fuel it into a platform and then this platform that I'm talking about is called Amplero, and it is an AI sort of recursive machine learning that is actively testing activation. So variables around tone, variables around imagery, various variables around length, calls to action, as well as channels, as well as sequencing. So all of the activation variables that can make you feel like you're getting a very personalized and relevant marketing interaction.

Jim Marous:
I mean, the good news is, what you're saying is, an organization doesn't have to have perfect data in order to be able to apply it and do some really good use cases and early victories, correct?

Sarah Welch:
That's 100% right. Yes. And one of the things that we hear a lot is, "Yes, we know personalization's really important and there's a five year plan, a five year roadmap for us to get to some mythical perfect place of personalization." And the fact of the matter is a lot of that work tends to be personalization efforts that are built on fundamentally linear technology platforms that are basically trying to use humans to do the testing and learning that will drive that dynamic empathy, if you will, that really isn't where humans are best applied. It will take a lot longer and be a lot more expensive trying to do that than by using machine learning and AI as a plugin to those platforms you already have to drive the dynamic experiences.

Jim Marous:
So you've talked about AI driven marketing. How is AI driven marketing execution different from the traditional rules based execution and what makes it better?

Sarah Welch:
Rules based has its place. When a treatment needs to be the same for everyone and needs to be predictable or unchanging, or when the clarity of communication is more important than personalized nuance, rules based makes a lot of sense, but... And rules space was the first logical step in an automation journey. It enabled you to have more always on versus episodic campaigns. But we are now in a world where rules based is sort of hitting a wall. What you need are recursive closed loop testing and learning to mimic that human banker exchange for example, that I gave earlier where you are learning and watching based on stimuli that you're putting out there and adapting and having machine learning drive the test cell set up, the variable sets that get delivered in a dynamic creative optimization payload. That is where machines are best applied.

Sarah Welch:
And then the marketers on the team can be really focused on the bigger strategic questions, which is what humans are great at, which are, what are the things that seem to be working? Why? What do we need to do to create better, more compelling marketing interactions for our customers, as opposed to the engineering of, for each micro segment that you may want to be reaching, what is the exact permutation or sequence of permutations that you need and then engineering those by hand in a platform. So the machine learning that automates that is really what enables marketers to make a dramatic leap forward in terms of personalization with the staff they have, not an army of 50, 60.

Jim Marous:
I talk a lot on this show about the fact that just developing these great models and figuring out where customers fit and giving the marketing in a great report is not really the power of AI and personalization. It's really along the line of actually deploying it and actually making it so that you can deploy these insights across a whole organization. I know your organization has written a lot about the role of branches going forward. And yesterday I wrote an article for The Financial Brand around technology enabled employees and how they can become more valuable, let's say, branch employees and let's talk onboarding, which you brought up earlier.

Jim Marous:
The ability to use branch based employees to take insights that are developed from your machine learning, AI learning models, and actually come up with solutions to say, "Here's what some of your customers should get next," that next best product, but then deploying the insights in a way that makes it so there's that human contact coming back. So let's say a customer opens account digitally, no organization has really mastered the cross-selling or the upselling engagement level of these digital accounts, so the reaching out by a human becomes very key. Do you see your tools that you develop at Curinos being ones that can actually be deployed across the organization to actually empower people beyond marketing with the tools that can make it so consumers actually benefit from better engagement?

Sarah Welch:
Absolutely. I think this is an area of huge untapped opportunity for banks, is how do you leverage technology to almost be like an iron man suit, if you will, for the front lines that you have and giving them stimuli and insights about the customers they're interacting with, or either in real time, so that they can interact with them in more appropriate and relevant ways. Or by engaging humans in outreach, as you said, at critical windows where it is really going to make a big difference in the overall value of that relationship. The interesting thing is that it's not going to be a one size fits all. There may be... So let's take a segment like high earners, not rich yet. That's a sexy segment. Everybody wants to win it for tomorrow because that's where the mass affluent population of tomorrow is today.

Sarah Welch:
If you think about, there may be micro segments in that population that would value a proactive outreach from a human banker at some point in the onboarding process. There may be a micro segment that wouldn't ever want to talk to a banker in-person. They prefer other channels, or to be totally self-directed. I think the promise of AI to some extent is to be able to learn which are the customers that would value an outreach and who, not just who would value that outreach, but who would because of that outreach have a deeper and higher CLV over time, a better relationship, a better quality relationship.

Jim Marous:
It's interesting, throughout all the discussions today, one theme that keeps on coming underneath the radar a little bit is your focus on test and learn. On the focus that every institution combined with every consumer is a different matrix you have to really learn over time. And I think we forget that. We take data, we think data's perfect and we push things out and we don't listen as well. And that's a part of marketing that at the end of the day, if we don't listen to the consumer, we're not going to be nearly as effective. And unfortunately, that's where you can't apply everything you've learned at one organization deployed against another organization because the whole dynamic of each company is different. The history of the company, what they focused on, what was important, their brand message, their mission statement, how empathetic they are in the community. So really if I'm not mistaken, a lot of what you're doing is not just a deployment of insights, but it's actually listening to how those insights are used and how we can get better over time, isn't it?

Sarah Welch:
An insight is only useful to the extent to which you can act upon it and learn from it. Yes, the activation, so the finding of the insights and then the activation of, against those insights in treatments, and then the learning from what's working, and what's not working in a closed loop iterative way is absolutely central to, we think, modern marketing and how to do it successfully. Jeff Bezos has certainly been known to talk a lot about testing and learning as what he thinks is one of the most valuable things and the biggest drivers of value for Amazon. It is effectively what we humans are doing at our best too. We are rapid test and learn machines. And so I think what the opportunity is for marketing to bring that into digital channels and digital applications at scale. And that's where we're not really even scratching the surface in most financial institutions today, but there's tremendous upside in getting it right.

Sarah Welch:
I'll give you an example of another use case. If you think about card utilization as a use case that a lot of banks think about and have marketing campaigns built against, you might say, "Okay, we have models that are running that sort of say, 'Okay, Jim is eligible for a campaign to drive card utilization. We've noticed he's fallen off over the past X amount of months or days. And so he's now eligible for that campaign.'" The marketing group is working and it's, I don't know, is it a three month, four month process to stand up a campaign for the entire population of cardholders that is eligible for a utilization campaign?

Sarah Welch:
There may be a whole lot of different types of cards and therefore different types of incentives and condition offers that you want to put out. Each one of those would be assigned to specific segments, so just setting up that as a campaign can take a long time. And then, they'll run that campaign. Maybe they get one a quarter, let's say, that would be a very high functioning marketing team. And yet there are a lot of questions left on the table because you're not testing and learning within that context beyond a very micro amount. Does the amount of decline in utilization make a difference? Does it impact the give, get logic for any individual? What give, get logic is most compelling for any individual? Is it gas and groceries and points? Is it something else? And what is that, how does that change the uptake of the incentive?

Sarah Welch:
Are there indications that how you wrap the creative drives a differential in how much incentive you need to give away? I mean, Jim, if I deliver a nudge that is funny, that tips you to put your card back at the top of the wallet, and I don't need to give you any incentive, how many others like you are out there that I'm able to save millions in incentives? Can I get out ahead of declining usage with proactive messaging? What might that look like? So these are all big questions that are not being asked. And they're not because it's, until this kind of time of AI, they just really, haven't been able to be operationalized in linear workflows.

Jim Marous:
We've talked about a lot here today and there's a big nut to crack. I mean, most organizations are the starting gates pretty much. Is this something, is engagement with your firm something that is really only for the big guys, or can smaller organizations also scale in such a way that they can get benefits tomorrow from doing what you've talked about today?

Sarah Welch:
Yeah. I would say, there are definitely... Sorry, I'll start over. So yes, I would say that institutions across the board can take advantage of this kind of learning. Obviously machine learning likes data, so those with data scale will move faster with the learning cycles. However, there's still a lot of learning and optimization that can be done on smaller customer bases.

Jim Marous:
So finally, if you have one piece advice to give to our listeners who are keen to take action on improving their marketing process at their organization, what would it be? Where would they start?

Sarah Welch:
I would pick a use case that is the most important use case. So for example, is acquisition your primary imperative? What percentage of your acquisition dollars will be wasted because you have a leaky onboarding bucket. Pick your use case and then I am sure that within your organization, there is no shortage of hypotheses about how you could improve marketing interactions and how you would engage or want to engage with them in a more dynamic way during that critical window. Even just going through that thought exercise can help you, even if you don't have an AI platform at your disposal to go drive that dynamism, be able to materially improve your workflow.

Sarah Welch:
I mean, I think the other piece is, if we are moving from mad men to math men where really marketing is becoming much more scientific, organizing your marketing operations so that you are following some kind of routine test, learn, iterate cycle, whether you start on a quarterly basis or a monthly basis having a learning agenda and a test agenda and driving that sort of continuous loop and getting in the habit of that approach to scientific marketing is definitely a really good place to start.

Jim Marous:
Sarah, it's been great to have you on the show today. I think, you've really illustrated something that we keep on having as a recurring theme on the show which is, you can't wait until tomorrow for something that needs to be done today. Number two, you can't do it yourself. You need to partner, and you need to partner with an organization that's going to run at a faster speed than you are without you having to hold the leash the whole time, because you're not going to be able to move forward if you don't partner with an organization that's going to move much faster than you. You also are going to get the tremendous benefit of resource that you would never be able to acquire for yourself from these partners that have these resources internally and have the knowledge of multiple financial institutions that you work with that allow you to say, "Guys, we've seen this one before, this is not going to work." And it's in everybody's best interests as a final note, which you just brought up, to get that quick win, because that helps get funding for future wins. Everybody wins in that.

Jim Marous:
Your partnering in the solution, Curinos benefits. You're going to benefit because you're going to get budgets and you're going to move forward even faster going forward. This is not impossible. It's a fact of saying you just got to get unstuck. And this is really difficult because it seems massive. And I think the kind of relationship you talked about, Sarah, around how you break this down into, let's get these little wins because that makes it all so you get excited about the next one and yes, it's a big issue to deal with, but if you do it one step at a time, you're going to get victories and victories across the organization, which is really important. So again, Sarah, thank you so much for being on the show today. I really appreciate it.

Sarah Welch:
Thank you so much for having me.

Jim Marous:
Yeah. And one last thing. How do people get ahold of you and how do people reach out to your firm?

Sarah Welch:
So you can find us at curinos.com, C-U-R-I-N-O-S.com and I am on LinkedIn as Sarah Welch.

Jim Marous:
Great. Thank you again.

Sarah Welch:
Thanks so much for having me.

Jim Marous:
Thanks for listening to Banking Transformed Solutions podcast, a new podcast that focuses on innovative solutions for financial institutions. We'd like to thank Curinos, the sponsor of today's show. If you're a solution provider, why not discuss how you can help out bankers and credit unions solve a major marketplace challenge. Drop us in line. We'd love to hear from you. This has been a production of Evergreen Podcast. A special thank you to our producer, Leah Longbrake, audio engineer, Sean Rule-Hoffman and video producer, Will Pritts. I'm your host, Jim Marous. Until next time remember, amazing things are possible when you listen to the consumer and take action on their needs.