All lenders are constantly seeking ways to reduce risk. Whether it be through technology or better training for their personnel or some combination of both. The ones who have the most data on past or present clients have an advantage. Machine learning and artificial intelligence have been used for a number of years in lending to calculate risk in credit models among many other reasons. There are still manual processes in business lending underwriting but predictive behavioral technology could be the next level of risk-reducing tools a lending institution uses to its advantage to minimize default rates.
Background
First arriving from one company in 2019, behavioral voice analytics products are built on technology that has existed for years but the enhanced features are now gaining traction in the lending industry to enable lenders to reduce risk by analyzing the voice and speech patterns of a customer. Voice analytics technology has mostly been used in call centers to interpret the level of satisfaction a customer had on a call by measuring tone and other data points, but this predictive behavioral technology has found more use cases such as in predicting loan defaults in lending, in human resources, sales, and collections.
Who offers this software
The claim by one of the software providers called Voicesense is that by combining signal processing, psychology, and machine learning with an algorithm, you can predict to a certain level of accuracy that a loan applicant will default. Interestingly the software does not interpret the words or content of a conversation, instead, it examines the pace, intonation, and more than 200 prosodic parameters of a person’s speech in seconds. It connects speech patterns to types of behaviors that would typically follow. A complete personality profile is sent to your CRM in seconds for analysis in real-time or can analyze post-conversation.
There will be many other applications as the technology gains traction and proves its benefits to allow companies to improve their operations and bottom line. Voicesense indicates they provide businesses with clear-cut conclusions like whether a customer will buy. This information could be an ideal solution to enhance your sales team operation or if not used appropriately might be data just taking up time and resources.
Voicesense, an Israeli company, already has clients in the alternative business lending industry and they are working on more. They developed some of the software by working in Israeli government intelligence and major corporations. Their Risk Management product is described as helping a funder make smarter loan decisions and improved risk management. They understand the manual tasks that underwriting teams go through but want to be able to provide you with more actionable information at the top of the application process. If you get a Voicesense report early in the underwriting process, you then would be able to adjust the requirements for a certain applicant or decline them altogether. This is similar to finding out about something in a tax return and then asking for more stipulations to compensate for the increased risk.
According to the Voicesense website, “The VoiceSense Loan Default Predictor assesses a customer’s risk stratification based on different behavioral tendencies – risk taking tendency, impulsiveness, personal integrity, conscientiousness, rule abiding tendencies, coping, well-being and more.”
In their Collections product, they note that Collection agencies can enhance their performance dramatically by ignoring particular statements and focusing on a customer's emotional behavior and interaction patterns. The voice data collected can be used to determine who is most likely to pay on their past-due debt.
Summary
These products by Voicesense and others are very enticing. For a lender, if you could save ten percent on every million dollars worth of loans then the return on investment makes it worthwhile. The technology and products will only get better from here. Voice analytics will grow as the Internet of Things(IoT) expands into more products providing an immense amount of voice data to integrate into machine learning algorithms like the ones used in behavioral voice analytics.
To learn more about Voicesense, click below to set up a call.
*Funder Intel is an affiliate partner with Voicesense.