How banks are using customer data for personalized experiences
- Banks are increasingly mining data to inform customer outreach.
- The challenge with data mining is ensuring the outcomes offer value to the consumer.
If you’re a Capital One customer, an alert from the bank no longer just means there could be fraud on your account. It could actually mean the bank is trying to tell you that your cable company has increased the charges — and your bill has gone up.
It’s all part of a trend towards personalizing the customer experience. Joseph Whitchurch, Capital One’s head of customer experience and innovation for small business cards, said the idea to offer customers insights on their money habits grew out of fraud detection capabilities that the bank used to track transactions.
“We thought ‘what if we took that system and used it to solve problems for customers, like getting overcharged?’” he said, speaking at an industry conference in Boston on Tuesday.
It’s this thinking that led to the launch of Capital One’s Second Look program, a service the bank offers to monitor customer spending habits. It can give detailed insights into expense patterns of customers, beamed through push notifications. Some examples of alerts include being charged twice for the same expense, if a recurring charge increased, or if the customer has been extra generous with restaurant tips. The system is dependent on machine-learning algorithms, but customers have a say on the kinds of alerts they want to receive.
“[I realized that I] shouldn’t build a model first, but I should just test it and let my customers tell me how to train my models and let me work for them,” he said.
As data analytics technologies advance, banks and personal finance startups are using customer behavior data to tailor product recommendations or offer insights into customers’ money habits. Thanks to machine-learning technology, they are able to crawl through transaction data to detect spending patterns. They’re also looking at how customers engage with the content that’s presented — for example, how long they spend reading a blog post — to figure out how to improve the customer experience.
Other banks are trying to do something similar. JPMorgan Chase, Wells Fargo and Citi are reportedly mining customer insights to compete with startups that use customer data to offer product recommendations and advice.
Personal finance platform MoneyLion displays electronic “tips and tricks” cards and blog posts for its customers, but the content that shows up depends on customers’ activities, including their money habits and how they interact with the app.
“We have bank transaction data, credit behavior and location data; we want to be able to match that with a set of advice and recommendations,” said Tim Hong, chief marketing officer at MoneyLion, which connects to customers’ banks accounts through an API. Hong said algorithms are able to figure out when customers get paid, what they spend their money on, and recurring expenses like subscriptions. Based on these patterns, it makes product recommendations.
“It can range from ‘this is your cable bill and it’s 200 dollars, so you can cut the cord, and if you did, what are some of the options?’” he said. “We try to make these [recommendations] as actionable as possible basing them on the spending patterns and how you might compare to the population at large.”
Elevate Credit, an online lender that focuses on non-prime borrowers, also monitors customer behavior data. If users make payments on time, it offers them lower interest rates or credit limit increases. Eric Van Dohlen, Elevate’s chief analytics officer, said customer data mining is an area where startups have gained an early lead.
“[For fintech companies,] it may seem easier because a lot of that is how these businesses were set up; there are a lot of fintech companies out there that offer a better value proposition [relative to banks],” he said. “The banks have always had an interest in it, but their business models haven’t always supported looking deeply into the data to distinguish one kind of credit compared to another — they’ve always focused on a minimum credit quality.”
But despite advances in machine learning and data analytics, it’s still possible to give analysis and recommendations that actually can turn off the customer. As a result, companies have to be careful to balance positive and negative information about a customer’s spending habits, to ensure that the customer stays motivated to make better finance decisions.
Colin Kennedy, chief revenue officer at personal finance app Clarity Money, said it’s important to gradually give a customer recommendations and insights, in order for them to build confidence in their ability to save.
“Nearly all of us get paralyzed, scared and overwhelmed when we look at anything financial and it’s important to give the customer a chance to breathe,” he said. “As opposed to throwing 12 or 13 products, over time we give recommendations for products that can save users money. In this context, we are able to dig into the details more.”