When every issue is considered a technical issue
Tower Loan is a financial institution that offers personal loans, sales financing, and mortgage lending services across the southern United States. In 2021, online lending accounted for just 2% of their total lending. As part of a digital transformation initiative to shift more business online, the company set a goal to increase online lending to 15% within five years.
As this effort got underway, the company quickly realized they would need to make significant user experience improvements to make the online application process more intuitive for their customers. Tower Loan turned to LogRocket to surface frustration points in the application process and identify where and why customers were failing to complete applications.
“Having LogRocket has helped us identify and fix things faster. Before, we might know if something was wrong, but where was it? You see a few of these sessions, you narrow it down, and all of a sudden you’ve got it.”
In doing so, the Tower Loan team also realized they could use LogRocket to detect and prevent fraudulent loan applications, improving the rate of return on the money they loaned out. After just six months with LogRocket, Tower Loan’s online lending is at a record high and they’re identifying and eliminating thousands of fraudulent applications each month.
Opening a can of application worms
Director of Online Lending David Michel joined Tower Loan with a mandate to advance the company’s digital transformation. “My role is to increase the online lending channel at Tower,” he says. “It’s less than 2% of the total lending volume, but we want to get it to 15% within the next five years.”
One of David’s first initiatives was to conduct a full assessment of Tower’s online application process using Google Analytics and the help of external consultants. Right off the bat, they were able to identify a handful of issues causing customer dropoff. Fixing those led to an immediate improvement in conversion.
External consultants are expensive, however, and Google Analytics only provided surface-level information on user behavior. “It just shows the high level,” says David. “It doesn’t show the inner workings of the application. What does an applicant do? What does a visitor do?”
“I said, ‘We need to start finding out where else we’re missing stuff. We probably just got lucky on this one. That’s when we turned to LogRocket.”
In order to find the next set of issues, the Tower Loan team knew that they would need a deeper understanding of user behavior. Tower turned to LogRocket to see the user experience through the eyes of their customers, looking to use empathy to build a better loan application process.
Streamlining the loan process
Tower’s initial goal with LogRocket was to identify opportunities to streamline their online loan application process. After implementing LogRocket, David and his team immediately pinpointed areas of friction in Tower’s online portal. In one instance, David recalls, “the design of the application didn’t make sense. The text and the color of the font looked like links, but they weren’t.
In another example, customers were sent to DocuSign to complete their application. “After they went to DocuSign,” David says, “they were bounced back into the application, and they landed on a page that said ‘Your application is incomplete.’ The customers were at a loss for what to do and would often start a new application.”
“You get to see what actual applicants are doing, as opposed to knowing what to do and trying to pretend you don’t know what you’re doing. We get to see customers’ real keystrokes, their understanding of what the app is.”
Making these identified fixes was also easy, as LogRocket provides a point of connection between David and Tower Loan’s engineering team. When a new issue is identified, “I copy the session link and send it to our developers,” says David.
Because both David and Tower’s developers see exactly what the customer saw, there are no miscommunications around what went wrong, and the developers understand exactly what they need to do to remedy the problem. “We close the loop and do something to prevent it from happening again going forward.”
Since implementing LogRocket and making UX improvements based on the issues they identified, Tower Loan has seen a significant increase in conversion rate, leading to their best month ever for online lending in May 2022.
Nipping fraud in the bud
While LogRocket was initially brought in to improve conversion rates, Tower Loan has discovered other ways to leverage its capabilities to help the business. Watching sessions one day, David realized that LogRocket could be used to battle another of Tower’s ongoing issues: fraud.
Tower uses multiple tools and services to detect and prevent fraud, but they fall short in some areas. “They are great at identity theft fraud,” David says, “but not technical fraud.” Many of the fraudsters in the latter category are looking to take out loans without any intention of paying the lender back.
“They may not intend to pay you back, but they don’t do anything to trip the fraud detectors,” he says. “They don’t care about their credit. Those are tougher to get.”
“I use LogRocket to say, ‘OK, what are they doing?’ Any unusual behavior, we can flag it and use an API call to put them on a block list.”
David and his team use LogRocket filters — like IP address or event count — to identify potentially fraudulent sessions, then use session replay to confirm the attempted fraud.
“You get three applications from the same IP address within five minutes, and they’re using prepaid phone services. That’s unusual,” says David. “Or anything with over 800 events in a single session — let’s look at those. I can go and watch somebody fire off [loan application after loan application] until they finally get something that goes through or they give up.”
When they see irregular behavior in LogRocket, Tower Loan can make an API call to one of their fraud prevention services to confirm their suspicions, then put the fraudster on a block list to prevent any current or future applications from being approved.
Through this approach to fraud prevention, Tower Loan has reduced the number of fraudulent applications by over 20%, which amounts to thousands of fraud attempts prevented every month. On top of this, online lending amounts are at an all-time high.
Differentiating between fraud and confusion
Sometimes irregular behavior isn’t a sign of fraud — it’s another instance of frustration within the application process. “Customers would apply twice within a few minutes or some other unusual behavior,” says David. “First, red flags went up to see if it’s attempted fraud. But that wasn’t the case. Customers were in the application, and it was giving them ambiguous text. They didn’t know what to do, so they inadvertently started a new application.”
In instances like this, Tower Loan’s use of LogRocket comes full circle. They can continually monitor loan applications to proactively prevent fraud, and at the same time, they can identify new opportunities to improve the loan application experience.This combination is helping them to achieve their goal of making online lending 15% of all lending by 2026.
Tower Loan is also in the early stage of expanding their use of LogRocket to monitor the effectiveness of their spending on advertising and partnerships. By combining UTM parameters and LogRocket conversion funnels, Tower can measure the success rates of different inbound channels to better understand how to allocate marketing spend.
“When someone hits the site from a Google ad or an email campaign, we’re looking to LogRocket to see who doesn’t complete the process and why they drop off.”
LogRocket also helps identify redundancies in Tower’s partnerships. After seeing a conversion rate significantly below average for customers being referred from one of their partners, David and his team went to LogRocket to investigate.
When the only issues LogRocket turned up were also affecting all other customers, David reached out to the partner to see what was going on. “We learned that they’re filling out an application with the partner, and then they come to Tower Loan and they have to fill out the whole application again,” says David.
With this information in hand, Tower is now working with the partner to pass more information when a customer is referred, streamlining the application process. As the process changes, Tower will be using LogRocket to track the impact of the changes and identify any issues that prevent the experience from being as smooth as possible.
Want to see how LogRocket can improve your conversion flows? sign up for a free 14-day trial.