Case Studies
90-Year-Old Credit Union Reduces Average Handle Time 66% in First 90 Days of Pindrop Implementation
Using Pindrop® Protect and Pindrop® Passport, a 90-year-old credit union with 7.3B in assets, reduced Average Handle Time by 66% (from 0:54 seconds to 0:18 seconds) in 90 days. Pindrop® solutions streamlined the call handling process, enabling agents to quickly identify and authenticate callers, reducing the time spent on each call.
Before integrating Pindrop® solutions, the credit union suspected that they were being targeted with a high volume of fraud from a few bad actors. By how much? They couldn’t say, and had no clear path for uncovering the suspected fraud. With the implementation of Pindrop® Passport and Protect, our solutions helped this 90-year-old credit union identify repeat fraudsters who were running targeted campaigns and enhance the experience for their members by reducing average handle time (AHT) by 66% (from 54 seconds to 18 seconds) in the first two months of implementation.
About the Credit Union
Established in 1935, the credit union is a member-owned financial cooperative with $7.3B in assets, 260k members, and 920 employees across 60 locations.
The credit union is unique in that it serves both personal (B2C) and business (B2B) banking members, adding a layer of complexity to their fraud risk management strategy. This dual focus necessitated a sophisticated solution capable of mitigating financial risk for both consumer and commercial services.
Member Challenge
Like many financial institutions today, this credit union was proactive in addressing the rising fraud. For example, in 2023, the fraud call rate for banks was 1 in 699 calls vs. 1 in 976 calls for credit unions. The fraud call rate for credit unions rose 71% between 2021 -23.
The organization served both business (B2B) and consumer (B2C) members, each with a unique set of risks to accommodate. With B2B members, the credit union faced challenges typically associated with retail banking, like identity theft, card fraud, and phishing attacks targeting individual members. These risks often involve high-volume, lower-value transactions, requiring comprehensive identity verification processes and real-time transaction monitoring. Additionally, consumer-specific fraud prevention should balance security with user experience, as personal banking customers expect seamless, frictionless interactions.
Conversely, B2B members introduce a different risk profile, as seen in larger transactions, like complex account structures, and more sophisticated fraud, like vendor payment fraud. Business accounts often involve multiple users with varying levels of authority, requiring more nuanced access controls and approvals. Additionally, commercial members may require specialized services like wire transfers or ACH payments, which present a new and unique set of fraud vulnerabilities.
The intersection of these B2C and B2B risk landscapes created a challenging environment for the credit union, where a ‘one-size-fits-all’ approach to fraud prevention was insufficient.
Member Goals
For the implementation of a new solution, the credit union had three clear goals in mind:
- Reduce average handle time for call center agents and improve call friction (time spent in validation) for members
- Detect and ultimately reduce fraud throughout the call center
This bank takes responsibility for security across its three business segments – retail banking, commercial banking, and wealth management with the highest level of priority. The bank’s thousands of colleagues focus on helping both individuals and businesses be successful in operating across its markets in Texas, Arizona, California, Florida, and Michigan.
Before Pindrop® Solutions
Account takeovers were on the rise for this credit union. For example, potential fraudsters were repeatedly calling in using spoofed numbers. As a first step, the credit union wanted to understand if a repeat caller was targeting them, and second, they wanted to detect fraud when it was happening. The credit union’s Assistant VP of Digital Services shared, “We had one fraudster who called in requesting a line of credit for $30,000. They did this by moving the money out of a member account and into a shell account using the member’s information. Bad loans were also rampant because members easily bypassed the Knowledge-Based Authentication questions (KBAs). And, this was just the fraud that we knew about.”
Although fraud was rising, the credit union was determined to find the right solution and justify the cost. However, they knew that call center agents were uneasy, acting as the front line for fraud, and they were losing agents.
The credit union was aware of the evolving fraud landscape, where the human ear can no longer be relied on to detect a synthetic voice. With the proliferation of generative AI, easy accessibility to AI voice engines (+800% increase from ‘23 to ‘21)1, and low cost to use, the way we see and hear content has irrevocably changed. Today, the ability to manipulate audio is easy, and humans are unable to tell the difference. For example, in a study by the International Security Program, human detection of AI generated audio was labeled ‘as good as a coin toss’ , detecting fake audio with just 54.5% accuracy (compared to 62% accuracy for fake images). Traditional methods used by organizations like KBAs and one-time passwords (OTPs) are also vulnerable to synthetic audio attacks.
Why Pindrop
“The trust factor was there”
The credit union was seeking possible solutions for about a year, with a primary goal of ensuring security was ‘job zero.’ The total time needed to implement a solution was also an important consideration; how quickly could they mitigate fraud while doing so accurately and securely?
According to the credit union, others in the space offered various fraud detection features, but none had the entire package that Pindrop offered. Additionally, internal advocates within the team who understood the fraud landscape (e.g., the security team) advocated strongly for the solution and knew it was needed to help the organization uncover and mitigate fraud. The Quality Assurance Manager added, “Our Cyber Security Manager was a strong internal advocate for Pindrop. He said that Pindrop had the best technology vs any other provider.”
After seeing a demo, the credit union met with several other vendors who needed more member references than Pindrop. A consistent message they heard from existing Pindrop customers was, ‘Pindrop saved us.’
“Pindrop created the original technology, and it was easy to build a trust factor with the Pindrop team; they were the experts,” said the Quality Assurance Manager.
Ultimately, the team was approved to proceed with Pindrop because they demonstrated the product’s ROA.
The credit union went with Pindrop because of our:
- Strong reputation
- Premium technology
- Expertise across the team
What ROA did Pindrop® Solutions deliver?
“Once we implemented the Pindrop Passport and Pindrop Protect solutions, things took off.”
Since deploying Pindrop’s fully integrated authentication and fraud detection solution, the credit union has significantly reduced the success rate of fraud within its phone channel, seeing a 50% reduction in the number of fraud incidents overall.
Launched in September, 2023
Once implementation was complete, the credit union was surprised to see how many fraud attempts were happening; it was considerably more than they initially thought.
When Pindrop® Protect and Pindrop® Passport were turned on, credit union members were initially apprehensive; they no longer had significant hurdles, such as answering lengthy questions, to clear when getting through to the credit union. For example, within the first two months, the credit union went from an average of 54 seconds to identify/ validate a member, down to 18 seconds on average. Member satisfaction improved significantly, and average handle time was reduced, which meant average wait time also decreased. In fact, the credit union observed a 67% decrease (36 seconds) in time to authenticate members.
Additionally, the credit union’s agent experience improved dramatically, reducing overall agent attrition in just 90 days. For example, after launching Pindrop Protect and Pindrop Passport in September, the credit union’s employee turnover was the lowest they had ever seen in December, just one quarter after implementation.
Since the deployment of Pindrop Protect, the credit union has experienced an 88% reduction in fraud incidents occurring within their call center channel. In one month alone, Pindrop Protect detected nine fraudulent calls against an alert rate of 0.68% and a review ratio 241:1.
Since the deployment of Pindrop Passport, the credit union authenticated 23,188 calls, and low-risk authenticated an additional 11,132 calls. In one month, Pindrop Passport demonstrated an increase of 1.54% in enrollment penetration (0:6min), which measures the percentage of calls tied to previously enrolled identity claim vs the call population with an identity claim.
Not Just ‘Another Vendor’
Don’t take our word for it, the Credit Union shared their experience with the Pindrop implementation as well as their experience with our team:
“The ongoing support from the Pindrop team has been outstanding- they are extremely easy to work with, and they always deliver on their word. The service provided is white glove and turnkey. They are by far one of the most flexible and understanding vendors that we have worked with. Their team took the time to ensure that all stakeholders were armed with the information necessary to ensure that we could hit the ground running. Throughout the entire process, they answered our questions thoroughly.”
“In fact, Pindrop makes us aware of additional features that we might not be taking advantage of that are included based on the contract we have, like Custom Attributes. That is not something that you would expect from a vendor. This kind of service is what truly sets Pindrop apart from others.”
– Assistant VP, Digital Services at Credit Union
Sources Cited
1. Open source projects on Hugging Face leveraging an AI voice tool (from 38 in ‘21 to 352 in ‘23)