Capitalizing on the Hyper-Personalized Future of Financial Services

The urgency of personalization and customer experience over the next five years is being driven by customers that are demanding products and services tailored—in real time—to their individual needs and smoother, frictionless experiences across platforms and channels. The ability of banks to deliver on both will drive customer acquisition and growth. Artificial intelligence is playing a central role in banks' ability to use data-driven insights to capitalize on the hyper-personalization of financial services, including predictive engagement for real-time CX. Yet the vast opportunities must be balanced with operational commitment and execution on privacy, trust and compliance concerns.

Transcription:

Holly Sraeel (00:11):

Hi, I'm Holly Sraeel. I'm the Senior Vice President of Live Media for American Banker. We're here to talk about capitalizing on the hyper-personalized future of financial services. To my far left is Erin Crawford, Vice President, Head of Consumer Digital-Payments and Money Management for Fifth Third Bank. And to my immediate left is Rab Govil, Founder and CEO of Naehas. Alright guys, let's go. Let's dig in.

Rab Govil (00:36):

Let's do it.

Holly Sraeel (00:37):

Let's do it.

Rab Govil (00:38):

Wow.

Holly Sraeel (00:42):

So Erin, we have talked in the past about how banks can strike the right balance between delivering hyper-personalized experiences and maintaining customer trust and loyalty. Can you talk to me a little bit about that?

Erin Crawford (00:55):

Yeah. Oh wow.

Rab Govil (00:57):

That's good. I like it.

Erin Crawford (00:58):

I like it too. Actually. I can hear myself a bit better than everybody. First of all, thanks for joining us today. I know everyone's probably getting pretty hungry and just so everyone knows, we know each other a little bit and do some business and he Nehouse offers our, is our offering system at the third bank. So that's why we're up here with you today. But yeah, as we talk about hyper-personalization, we often talk about how much is too much, how much we don't want to target the customer too much. And as we think about from in the money movement and financial wellness spaces, there's often conversations around how much of my data is too much of my data. When you talk about receipt level data or office spend, and as people grow their financial wellness capabilities, there's almost a creepy factor at some stage.

(01:50):

And so there's a trust line there that's like, I want you to be able to leverage my data to know things about me that help me, but I don't necessarily need you to know that I bought this one item at the grocery store or online at a receipt level. And so I think there's a super hyper hyperfocus on balance right now as we're moving into sort of an open banking era. And I know that there's some regulations that have changed or been kicked down the road a little bit. And so for us, we focus on the things that meet our customer's needs immediately. One thing we were successfully able to do, you could say hyper-personalization might come in the form of an offer and we'll talk about some of maybe the hyper segmentation and how you target people based on different neighborhoods in a little bit, I think.

(02:43):

But one area of focus that we have said, how can we really focus on what the customer needs to do is in our onboarding journey, we launched in our mobile apps in 2023 and we've focused on making sure they are very specific on the things they need to do based on whether they're a preferred customer with us or a everyday momentum banking customer. We've said, we're going to give you the exact things you need to sort of get in, get set up and get going with your account. And that has, I think one of the areas where we've seen a hyper focus and personalization on meeting the customer's needs to get them up and working. That I might say has been a pretty good success for us.

Rab Govil (03:28):

I think that's the idea behind balancing the use of data and then truly delivering value to the customer is an important balance. I remember very, very early, this was almost a couple of decades ago, I was meeting with Lowe's, the home improvement people, and this marketing manager comes into this room and he goes, this person just bought some lumber and then I saw they're buying some nails and they're thinking of buying this thing. So I think they're building a deck. So I know this person's building a deck, so I'm going to reach out and tell them they're building a deck and how can I help it? And I said, no, don't do that. Please. That be really, really, really scary for somebody to hear that so much about it. So I think over the years as we have worked with a lot of financial services institutions, have we used our platform to do this level of hyper-personalization?

(04:26):

You need to have, for lack of a better word, some ground rules. And the best ground rule that we have picked is are you going to be truly helpful in what you're doing in making their life easier? If it's not truly helpful in making their life easier, then think again, if you really want to leverage the data in the way you're planning on data for using for hyper-personalization. And the rules are very different when they're an actual customer and versus where you're trying to acquire a new customer. So we are a platform provider that a lot of the institutions use to manage personalized products and offers both on the acquisition side as well as on the customer side. And reality is that you have enough data on both sides to get very, very personalized. It's not even acquisition through all the different bureau data. There's enormous amount of data available to be able to do that. But we specifically say this, be very careful about using the data on the acquisition side in terms of how you use personalization but offers and getting them to incense, something works really well, or creating products that are truly, truly personalized and they feel that, man, this product really works for me in the segment of the lifestyle they're in. And then when they become a customer driving more of the personalization using service, how do you make them useful to you is what generally ends up driving a lot of value across the chain.

Holly Sraeel (06:12):

I want to ask Aaron a question and then Rob you can respond to it, but Well, it's a two part question. One, how much personalization is too much personalization that might, I dunno, offend customers or bother them. And then you talked a lot in our prep call about using the data to tell you about the customers and their behavior. So you want to talk about how you balance those two things.

Erin Crawford (06:43):

I think how much is too much goes into if you stop seeing activity, it's been too much. And I think just recently just doing some competitive analysis. I've gone to a couple websites of our competitors recently and the first thing you see is a bunch of their products pages and I'm just trying to get to my credit card and when I authenticate, the first thing I saw was maybe my account balance in the top right corner and then the rest of the real estate is how I could upgrade my credit card and that's not what I'm there to do. And so that was a really eyeopening when we think about why am I being targeted for these other things? I might take an offer, but if I'm happy with the product, then I think one thing we've successfully done in our channels is revamped our customer recommendation engine to be able to say what's that next best offer? And not make sure that they're static, but that they are a little bit more targeted based on some of the activity that you're doing. And then what was, sorry it was hard to hear you Holly. What was your second question?

Holly Sraeel (07:56):

My second question was to, if you could speak a little bit about how closely you look at the data to inform what you do for what you serve up for customers in a personalized fashion?

Erin Crawford (08:08):

Yeah. Data's sort of at the forefront of everything that I do in my job. We look for all sorts of trends, whether it's with consumer trends, even within our vendors and how they're performing on if customers are making on-time payments. I would say one of the really interesting areas where we've leveraged data to maybe target or help customers is in our customer solution space for collections. There's a tremendous amount of data that goes into our collections groups holistically across the bank, and I think has been a good cross collaboration across our multiple lines of business, between lending deposit, the channels, retail, the call center, and we've used data across collections to make sure that we're enabling the next best offer to help them get out of delinquency. So what's the best hardship plan for them? How can we meet them where they are? Is there a settlement offer available to them because their behaviors have shown that they might not be able to be on a payment plan.

(09:21):

So I think that's one area where we've leveraged the data and because we are holistically from a financial wellness perspective, trying to meet you where you are, you're either behind and need to catch up and we're giving you the tools and the appropriate focus that you need to get caught up and get out of delinquency. You're living day to day or trying to manage a budget or you have a little bit more or a lot more and can try and save for the future. And one of the ways that we've sort of expanded into each of those spaces is through data. We just launched a free estate planning services for all our customers and we were able to do that because we had offered a product through trust and will that gave a discount on services for estate planning and trusts and changed the discount codes for a while.

(10:11):

And then we targeted a small group of our existing customers and said, would you take advantage of this offer if really this benefit if it was free and we saw a surge because we were able to break that barrier to entry through providing that service for the customers that maybe even if they are behind, you still can't put a price on your family and maybe your car is delinquent, but we still are trying to protect you holistically on your financial wellness journey. And so we've seen some success with the data in those types of stories to help our customers in those spaces.

Rab Govil (10:50):

Yeah, I think it's kind building a little bit on the last question is all your specific examples are, Hey, I'm actually trying to help the consumer do a better job managing their finances, right? That's right. That's where it works really, really, really well. So we have a little broader perspective. We work with a lot of these financial institutions and as part of this broader perspective, I would say where we are seeing the use of data, at least on the acquisition side, is starting to get to more specific segmentation that is just not FCO code driven. Initially, most of the segmentation was, Hey, I need to build up my product portfolio or I need to get deposits from a certain set of customers. So the main segmentation we are planning on doing is going to be basically by offer like FICO or something else of this sort.

(11:53):

So now what we are seeing with all our customers is they're starting to now do the segmentation a little bit more deeply because they understand is different segments will perform a lot better from either a net interest margin basis or they'll perform a lot better from a delinquency basis. So they're trying, but they still want other parts of the portfolio that might have a higher interest rate. So since they're trying to balance between what I call performance as well as getting the right amount of volume going through this. So I think on the acquisition side, we are starting to see much more deeper segmentation that we have seen over the last years. And that's why you're seeing these very specific targeted offers that have started to emerge. And in fact, we're seeing a lot more customers buying our software just to accomplish that level of targeting. But that's what we have seen on the customer side. I would build on what I heard from Erin is that we are seeing more service related type of offers like do this because you're trying to really incent the customer initially to engage with your products because that's probably are going to create more stickiness and if you create the right product, it makes a difference. I'll give you a specific example.

(13:23):

One of the top four banks, they were putting out these offers where somebody would sign up for a checking account right after somebody signed up for a checking account and they were basically saying, okay, I'm going to get $500. If you do these four things, you do a direct deposit, your average daily balance should be for this amount of time for so many days. But the problem was they were not using any of the data to engage the customer after they've signed up for the account. The only way the customer could actually find out where they were in the journey would to make a call to the call center and after they would make a call to the call center, a call center person would open up a ticket and say, we'll get back to you within a week. So you have just spent so much energy and time acquiring this customer, getting them to buy your product, and now you're not using data to inform them. Along the way is where they're in the process. So they put what is called on the online banking app and on their website on the authenticated part of the website, what they call a tracker, an offer tracker. And basically whenever they logged on or they got an email, the tracker would tell them using the data to say, Hey, you are 10% to your journey.

(14:57):

You did the direct deposit, but you're not put enough new money into the bank account to achieve their goal. Once they leverage the data to kind of enable that service experience, the number of calls to the call center went down by 60% and out of the calls that went to the call center, 44% of the disputes went in the favor of the bank because they could just use the data to provide precise information at the type the consumer needs to that. So those are super useful ways of using data in a personalized way that truly adds value to the bank, to the institution as well as the consumer because you want to do both ideally.

Erin Crawford (15:46):

Yeah. Remember when I was talking about that onboarding checklist?

Rab Govil (15:48):

Yes.

Holly Sraeel (15:49):

Yeah, they were saying the same thing.

Rab Govil (15:51):

Same

Erin Crawford (15:51):

Thing.

Holly Sraeel (15:53):

So what are the biggest operational challenges that banks face in implementing real time cross channel personalization? How can they overcome them?

Rab Govil (16:02):

Okay, so what are the real operational challenges of these cross channel optimization? How long do we have? I could talk about that for days and I'm sure Aaron will have this thing. Well, it's fascinating in all the implementations we have done, we just had a half day workshop at a top 10 financial institution just last week in Washington and we had the CIO there, the chief marketing officer was there and the head of their digital experience was there the first two hours. We just spent time to understand for everybody what an offer means. They couldn't agree internally what an offer means. One person thought that the offer was just incentive. The other person thought an offer meant the product attributes, the disclosures, the channels that goes into it. So I think one of the largest operational struggles that exist in most of the banking institutions is you don't have, first of all just clear definitions within the institution of what's an offer?

(17:29):

How do I get the offer agreed upon and approved? How do I get that offer into the hands of the consumer fast enough? And then how do I track all that stuff on the backend to make it end? The average time it takes somebody to get an offer out in the marketplace right now across our customers is 120 business days. So think about that, that I need to start working on the offer I'm going to deliver into the marketplace or a service. I'm going to create something I'm going to do in June. I need to start working in January. So I think, but we have seen our customers able to change their processes, change their mental models, do simpler things that they've been able to take that from one 20 days to less than 20 days to do that. But it's a lot of process issues, a lot of definitional issues, and obviously all the legacy issues.

Holly Sraeel (18:37):

Okay. Aaron?

Rab Govil (18:39):

Aaron, operational challenges. Yes,

Holly Sraeel (18:42):

Thank you. Well, first of all, Aaron wants to draw a distinction, if I remember correctly, between offer and benefit.

Rab Govil (18:49):

Yes.

Holly Sraeel (18:49):

But we could come back to that.

Erin Crawford (18:52):

Yeah. So how are we doing on time? I think the operational challenges, at least from experiences that I've had across channels, is making sure that everyone's speaking about it the same way

(19:09):

Across our retail branch network, across our genie intents. It was interesting when we did launch the trust and will benefit, I won't call it an offer, I'll call it a benefit and I'll tell you about that in a minute, but we got some feedback from different groups from private bank. They were on board, we have appropriate communications through them. But in our genie intents, we hadn't forced upgrade our mobile app until a couple days after we rolled out this new capability and we started seeing questions in the Eugenia intents about the service for trust and will and the free estate planning services. And so until we forced upgrade where we had an intent available for them and making sure that it was streamlined and consistent. So we did see the volume of inquiries prior to the intent being ready, which was interesting, and we solved it by having that intent be ready, but making sure that everybody knew at every touch point that you have a response for the customer and making sure that we're all speaking the same language, they are directed to us or directed to trust and will where it's appropriate. If they are having technical issues, obviously we want to make sure that that feed back loop is coming through and back to us in the appropriate way. So I think from an operational standpoint across the connectivity, that's just one example where we may have seen some challenges on making sure that just consistently we're having a streamlined message when we go out the door with a new capability or a new personalized experience for the customer,

Rab Govil (20:50):

Getting the channels to be consistent, that's a hard problem, right?

Erin Crawford (20:55):

Yes.

Rab Govil (20:56):

And especially the non-digital channels, right? Because in that case, you got to get actually the human to talk about it the same way, and that's even harder.

Erin Crawford (21:06):

I actually walked into a branch the weekend before we launched that and just to make sure, because in our working groups and in our communications up and down the company, we thought everybody knew about this relationship. And I walked into the branch and I said, are you excited for the big launch on Monday with trust and Will? And they said, I don't know what you're talking about. And I said, this is the biggest thing we have going on in the channels. But it also is, it was very humbling to know that because we're a large organization that not everything is everyone's most important priority. So making sure that you do have those talking points and the consistency across the channels is incredibly beneficial for a big launch like that.

Holly Sraeel (21:54):

One of the things that I think is potentially most riveting with the ability to personalize what you bring to customers is hypers segmentation. Because what customers need in one neighborhood could vary greatly from another neighborhood. So can you guys address how you go about that? How do you segment? Do you want to take that one?

Rab Govil (22:18):

All right, I'll take that one. So there are a couple of things with hypers segmentation. First of all, we also live in a regulated world. So one of the interesting things is that there have been issues reported where people had tried to put offers or products together by zip code and by doing things by zip code, they got themselves into trouble because they were not following the rules in terms of the, because zip code was not good enough granular detail. So generally the processes we have seen for hypers segmentation is that most institutions will have certain fields that they can never use for any sort of segmentation. They call them protected fields and they'll say, these fields we are not going to use for segmentation because if we use these fields for segmentation, we will run afoul against lending laws. And if they do use them, there's a governance process that the way you're using those fields is appropriate and it's not going to hit any compliance issues. So I think the important part is that us in the financial services industry have limited to about how much hyper segmentation can do compared to, let's say Amazon. Because if you provide a certain pricing for somebody versus another person, we might get ourselves into trouble.

(24:11):

But I still would say if let's say an Amazon is doing hyper-personalization or a airline is kind changing the pricing on a hour by hour basis, our industry is still pretty far behind, I would say is if they're at a nine or a 10, one out of 10, our industry is probably at a two or a three. So there's still a lot of performance available to do it. You need to just put the right processes in place that you're using hypers segmentation in such a way that doesn't run afoul of any of your other regulatory environment.

Erin Crawford (24:56):

It's interesting, I've noticed this while we're up here. I'm not sure if you all have noticed, but I think just the way that we think about sort of targeting is so vastly different. Yeah. You are targeting very specific things through your systems to say, do this action, get this thing. And we have at least two and a half million active users a month where we need to make sure that the product is right for them and the benefits that sit on that product are right for them. And then from a hyper segmentation standpoint to make sure that sort of, there's a gauge across each level of those products that fits their needs. So it's a little bit different when we think about hypers segmentation. I think as I think about all of our digital consumers and what they come and need to do every day versus how you might target someone on a specific offer to take action,

(25:52):

And it's just a little bit of maybe where we are together, but not disagreeing, but just to have a different point of view, is that more holistically from a hyper segmentation about how we're doing based on different neighborhoods. In my space, I don't think about zip codes. I think about do you have financial wellness education available to you? Are you following the things to make sure that your account is secure? Because it doesn't really matter where you're coming from, if the product itself holistically and the benefits we offer on that product fit your needs, and then other strategic tactics we can take to go deeper in a market on specific things, yes. But I think that's where maybe we have just different perspectives.

Rab Govil (26:42):

No, I think service is very different than acquisition in some cases, in terms of and in service. I think a lot of things come to the point of what do they need at that moment in time in terms of doing this. It could be a life event.

Erin Crawford (26:56):

That's right.

Rab Govil (26:58):

A kid just went to college that's going to require different services they might want to offer or they want to buy a house. So I think life events and other types of things. And by the way, on the digital side, you have such rich data because they might walk into a branch once a quarter, but they might interact with your app or the website daily. Daily, right? So amount of interaction and amount of information you can gather about what they're interested in by just testing certain ideas and see what they like and they click on.

Holly Sraeel (27:34):

That's right.

Rab Govil (27:35):

Makes a huge difference.

Holly Sraeel (27:37):

Alright guys, we're at time. So I want to open up the floor to questions for Aaron and Rob. Do we have any questions? I can't say with the lights,

Erin Crawford (27:46):

They might be too hungry.

Holly Sraeel (27:48):

No questions. Oh my God. It's all anybody can talk about.

Erin Crawford (27:51):

He's

Rab Govil (27:52):

Got a question.

Erin Crawford (27:52):

Yeah. You're talking about

Rab Govil (27:55):

Personalization, find that balance value. How do you

Erin Crawford (28:03):

Actually customer direct inputs and feedback versus data analytics, ai, et cetera in finding that? Yeah, I think you have to ask for it in some ways or else you're going to be hearing about it. So I think there are a couple of things we do to survey our customers. You also see some of those satisfaction type scores where it's like, did you like this experience? So we have integrated tools that help us determine after they've completed certain actions, was that good or bad and can you give us some open feedback? And then we're recently sort of shifting our strategy on some of that because we used to, prior to our new mobile app launch in 2022, we had more direct feedback that wasn't just app store feedback to the app stores, but feedback to us in the channels where maybe as something as simple as this transaction's wrong and I went to an ice cream shop and it's showing up as a gym. There's very small instances where that happens, but it does happen on those smaller level merchants and just being able to get the feedback to say thanks, and we're going to go fix that. We're looking at more mechanisms for our customers to give us that feedback. Another example would be through our Genie Chat capabilities, being able to take feedback that way and then use AI and intents and different components to synthesize that to help us identify consistent problems where there's maybe an area of opportunity. But I think that's how we handle those things.

Rab Govil (29:36):

We see the same thing across our customer base, but I have to tell you, some of the best programs that we have seen run at our customers have been in where there has been an amazing product manager or an amazing experience manager in the role. Somebody who actually not only looks at the synthesized data but goes out and talks to customers. So you always feel that data will drive these things, but some of our best folks have been folks who have been amazing experienced managers, and they will just get some insight by actually going and talking to customers and then they come back, Hey, I heard this thing and now I'm going to validate it with data. Instead of starting with data, they come up with hypothesis and then validate it with data instead of the other way around. We have found that to be a very, very effective way of building great programs.

Holly Sraeel (30:37):

One more question if anybody has one. Okay. I'd like to thank Aaron and Rob for their expertise and their sharing insights today. Please join me in thanking them. Thank you. Thank you.