Consumers don’t really want products, says Sathish Muthukrishnan, Ally Financial’s Chief Information, Data and Digital Officer. “They’re looking for you to solve a problem for them, and money is the most personal thing that they have.”
Muthukrishnan oversees a team of technologists who work to help the firm stay competitive in a financial services arena crowded with both new and established entrants — and stay in sync with those consumers trying to solve problems. Based in Detroit, Mich., Ally is an online-only bank that traces its roots back to 1919, when it was created as a division of General Motors to help auto dealers finance their inventory. It holds $192 billion in assets, and is one of the top 25 financial holding companies in the US.
We spoke recently about the company’s mobile experience, AI, and hyper-personalization with Muthukrishnan, who formerly served as an executive at Honeywell Aerospace and American Express. “We’ve made all the foundational investments, and now we can unlock all of the experiences for our consumers and keep pushing Ally forward and make it the foremost institution in the financial services field,” he says.
• • •
You recently reimagined your mobile experience into a single app interface. Can you tell me a little bit about what you were trying to solve when you did this, and how it turned out?
A few years ago, we had six different mobile apps supporting different products and businesses that we had, all the way from auto to bank products, investing, mortgage, et cetera. We had to move at an unprecedented pace so we could satisfy customer expectations. For that to happen, we could not have such a complicated ecosystem of apps. At the same time, we were also seeing customers overlapping with the different products we were offering. Customers on the bank side would open an investment account or a mortgage account. We launched this initiative called One Ally. Part of that punch line was: We now have a single app facing all of our customers across all business lines and products that we offer.
What were we trying to solve? I talked about one aspect of customer expectations, changing and evolving fast, especially during COVID. Multiple decades of digital transformation happened within those two years of COVID, because everybody was becoming digital.
Number two, you were starting to see a variety of cyber attacks from across the globe. Having a broad technology ecosystem opens you up for a broader landscape that you have to secure.
If you have six different apps and data dispersed across them, you’re probably not going to be able to meet the customer expectations, secure your technology landscape, or surface the experiences that your customers expect.
And number three, customers started to think, “You need to know who I am and why I’m coming here,” so data becomes critical. If you have six different apps and data dispersed across them, you’re probably not going to be able to meet the customer expectations, secure your technology landscape, or surface the experiences that your customers expect.
One Ally sounds like a really big lift. What teams are responsible for this transformation and what main obstacles did they face that they ultimately had to overcome?
One Ally was a big lift, not just for technology, but for the entire company. It started with us looking at the dispersed data sources, bringing it all together. We cloud-enabled it, so it was easy and future-proof. More than 75 percent of our applications are on the cloud; 95 percent of our data is on a single data warehouse sitting on the cloud. We also modernized our network. The reason I bring up the network is that it’s the highway which the data traverses through to create experiences for our customers. We completely shifted to a software-defined network which created the seamless interaction between our applications, data and customer experiences, which led to us being ranked as one of the top apps across the financial services industry.
I have to thank not just my technology team, but all of the businesses that took a step back and said, “With the transformation of this front end, our entire background processes and methodologies have to transform. How do I co-exist with the rest of the businesses instead of just focusing on my own?”
I’m interested in one of your key goals, delivering more personalized user experiences. Can you tell me what that looks like in a practical sense for your customers today?
Generally speaking, people don’t love banking products. It’s cumbersome, and it’s usually trying to solve a big problem. Through savings buckets, we allowed our consumers to create separate buckets and label them. For example, this is my emergency fund, my honeymoon fund, my wedding fund, my down payment for my car. What we found was that shift in personalizing what you’re saving for, instead of having one big bucket where all your money is going, allowed customers to focus. They saved twice the rate at which they were saving before. That personalization, that experience, makes them feel as if they are our only customer, and that contributes to north of 90 percent of the retention rate that we have with our customers.
Old personal finance books might reference the envelope strategy, and that’s not really possible anymore for most people. It’s like that, but digitized. For the customers who use that, are there predetermined buckets they can choose from? Is it completely custom?
We have over 5 million buckets created already to date, and they could personalize every bucket. They could call it their cat food fund or adopt a pet or donations. [Customers can select up to 30 buckets, and the app comes with some pre-populated options.] That’s where the power of technology comes to life.
Being an online-only bank, how does that impact your decisions around tech innovations?
It infuses excitement and inspiration for us knowing that everything we do revolves around technology. We willingly take on the responsibility of being influencers of business strategy.
The first phase is ideation, where the business can dream up ideas that they want to build. The second phase is elaboration. We take that idea and collectively elaborate it to find out how this idea would transfer to a capability facing our consumers.
In our ecosystem, we created what we call an Ally Technology Operating Model. (See below for an illustration.) The first phase is ideation, where the business can dream up ideas that they want to build. The second phase is elaboration. We take that idea and collectively elaborate it to find out how this idea would transfer to a capability facing our consumers. What are we trying to achieve?
Then, we execute, and clearly track the timeline and adjust on a daily basis. That level of end-to-end visibility for technology has allowed us to work and create a very deep partnership with our business partners, ultimately benefiting our customers and shareholders.
Once a product reaches execution and beyond, such as the One Ally mission, I’m curious what happens after that. For example, as a tech user, I really appreciate when I’m able to provide constructive feedback to a digital solution, and I eventually see iterative improvement over time. I know this tends to be easier with smaller products and smaller teams. With Ally being on the larger side—10+ million customers and 11,000+ employees—how do you manage to prioritize improvement and user personalization over time?
We have various avenues to collect input. Our Ally Technology Operating Model allows us to figure out what works best for the consumer. Obviously, that is still internal facing. We very closely monitor all of the unsolicited and solicited feedback and comments we receive from our consumers. A portion of my digital team looks at all of the comments that we see in the App Store or they call the customers, and we collect that feedback and then feed it back into our elaborate session, saying, “We are building this capability for this product and here is the feedback that we got.”
Before we launch it, we create mockups that look exactly like the product that’s going to the consumer, and we have what we call usability studies where we bring potential consumers in and have them go through that experience. Meanwhile, my team is watching them go through the experience with minimal influence. We can see where the customers are struggling and where there are friction points.
Because it’s fully digital, I have the ability to see where consumers are spending time within each of the feature flows digitally, so we capture that, and all of it feeds back into our product and feature roadmap.
Each of my build team blocks anywhere between five to 15 percent of each sprint cycle, which is two weeks, for enhancements. We all obviously have to continue solving for tech debt, but applications that are newer will spend more time introducing these experiences.
About a year and a half ago, you announced your Ally.ai platform. Tell me a bit about what this is, who uses it, how it’s evolved over time and how it impacts your customers.
One of my career highlights and proudest moments while at Ally over the last five years is our launch of the Ally.ai platform. With the advent of ChatGPT in 2022, people jumped into using the platform. We took a step back and said, “This technology is fast evolving, and there are no guidelines around using it.” Even though the technology was so powerful, there were so many friction points.
We leaned into our cloud strategy, building a software-defined network, and what we said was, “Let’s build an AI platform that will serve as the moat between Ally and all of these external, generative AI LLMs [large language models] that are popping up everywhere.”
Ally.ai allowed us to behave like a financial institution from day one. Because of all the configuration we did for the platform, it allowed us to understand the input and the output that is going to external large language models so we can better protect Ally. We were selective about who gets access to it. In the initial days, people shut down access to ChatGPT for the entire company, but because we had this layer that was separating the external LLMs from Ally, we were able to selectively allow access to these LLMs from an experimentation standpoint.
Our first use case was with our customer service reps, who, on an average, take about 10,000 calls a day, each call averaging between 12–15 minutes. At the end of every call, they summarize their entire conversation with their customer. Ally.ai was quickly able to capture the entire conversation, convert it into text, and then summarize it and put it in front of our customer care reps. We ran that experiment for a few weeks, and around 90 percent of customers agreed the summary was accurate. It allowed our reps to fully focus on the customers.
Another use case is on the marketing side. We have several avenues with which we teach financial literacy to our consumers. We have what we call our Conversationally platform, where we teach our consumers about different financial nuances. They might come to our website and try to understand, “How does a mortgage work? How do I save for a down payment? How do I increase my wealth?” Those articles tend to be long, and we saw that customers would drop off after reading one or two paragraphs. Because we’re a digital organization, we’re able to track all of that. We used Ally.ai to summarize the entire article and show it to the consumers. If they like it, they can click and read through the entire article. Engagement skyrocketed after that.
Our internal audit team generally audits the capabilities that we build, and before any audit, they build what they call a risk control matrix. That is, “Here is all of the audit input that we’re going to test, and here is the output that we’re going to see.” Now, Ally.ai understands the capability and summarizes and creates the risk matrix for them so they can quickly jump into auditing the application and go deeper into it, saving them weeks.
All the way from customer care to an internal-facing audit, Ally.ai is adding value, driving efficiency and making our internal employees more effective on a daily basis.
I believe that magic happens at the intersection of three things: Adding customer value and experience, creating a business impact, and using the right technology to do that.
It seems like Ally.ai really acts as a flywheel at many points within the organization. Zooming out, where do you anticipate personalization in financial services headed in the future? And how must tech teams stay on track?
I believe that magic happens at the intersection of three things: Adding customer value and experience, creating a business impact, and using the right technology to do that. When all of those three intersect, magic happens.
Our customer expectation is evolving, and they no longer see banks as traditional banks. They expect the same digital, friction-free experience they have outside of financial services. It’s our responsibility to meet that expectation. They’re not looking for products. They’re looking for you to solve a problem for them, and money is the most personal thing that they have. So how do you help solve the problem? Maybe help them pay off their loan early, save towards a down payment, meet their retirement goals or pay for their kids’ colleges. All of those translate into experiences that we can provide our consumers, and that’s only valid if we start to understand our customers at a very personal level.
