Managing consumer complaints for banks and lenders is a big deal, whether you’re looking to avoid regulatory risks, improve customer experience, reduce expenses, or all three.

So why is it so hard to do well?

We brought together three leaders who know about managing consumer complaints, from identification and capture, and asked them to share their tips.

The big takeaway - you need a human-led, AI-powered process to address complaints management effectively.

You can watch the full webinar or read on for the highlights.

Major Insights

Takeaway #1. Complaints are a people issue.

All our panelists emphasized the importance of having a robust process for identifying, resolving, and acting upon complaints. 

The key? Human interaction and training are essential for accurately identifying and addressing issues. 

"The beauty of complaints to me is that it's just so common sense. You have to listen to people. If you don't listen to people and heed their word and investigate if they do have a complaint, then those complaints escalate or they even turn into litigation," said Lee Brockett, CEO of Cascade365.

But that’s easy to say and hard to do, Brockett added.

Anand Joshi, former VP of Global Collections at American Express, agreed. “It’s a people-intensive operation. Companies have invested a lot of money and a lot of time into the entire process.”

‍Takeaway #2. Complaint capture and identification is tricky, but central.

It almost goes without saying that if complaints aren’t identified and captured correctly, they can’t be dealt with.

Joshi emphasized that the identification process needs to be clear. “You need to have well-defined parameters set up so you can clearly identify that this is a complaint.”

He stressed the importance of having compliance experts as arbiters, saying, "Don't let the people who are on the floor design or decide what a complaint is. That has to come from somewhere else, especially from a legal compliance standpoint."

Scott Hamilton, Banking Strategy Leader at Prodigal, who previously worked at Capital One, Bank of America, and JPMorgan Chase, pointed out the logistical issues that make this so tricky. 

“The larger firms are asking a distributed network or thousands of agents to form judgments on every interaction they have as to whether there was a complaint in there, and they’re also asking those frontline agents to de-escalate and remediate in the moment.”

That's a lot to ask anyone to do and expect them to get it right.

‍Takeaway #3. Supporting people with technology is the key to effective complaints handling.

“The industry is evolving to have technology do most of the heavy lifting and then have humans go back in and interpret, strategize, and perform QA on the back end,” Brockett said.

Speaking from a large, mega-lender perspective, Joshi discussed the difficulty of scaling. “We would have thousands and thousands of complaints every month coming from multiple sources. So how do you handle these volumes with accuracy and precision?”

Hamilton underscored the opportunities of modern technology to support the vital initial steps of management.

"It's the front half of the process. It's identifying the complaints across multiple channels and aggregating them in a way to yield clean data so that you can then go after the tallest bar and issue quickly."

Those initial stages are where most large banks have received either MRAs or Consent Orders, Hamilton explained. But previous efforts to support efforts with technology have been a huge challenge due to low accuracy rates. Those failures have led lenders to return the responsibility to front-line agents and the problems with human accuracy.

‍Takeaway #4: AI trained in consumer finance helps where older technology failed.‍

Because previous solutions tried to address consumer complaints through faulty transcription and keyword and phrase searches, they were bound to fail.

But AI has the power to change all that. By using models trained specifically on consumer finance and moving beyond basic speech-to-text to incorporate sentiment and context, AI can deliver better than 90% accuracy.

That means instead of relying on already overloaded customer service representatives to try to manage all the steps of the process on the fly, AI-powered solutions can handle those tricky first pieces, leaving agents to manage the personal angle.

“We use highly trained models to take that in-the-moment judgment off the table and automate the identification, the holding the hands of the agent to de-escalate or remediate, the note capture, and insights and QA activity on the back end,” Hamilton said.

Our panel agreed that this will always be a people-centered process, because complaints are about people. But the opportunity for banks and lenders to curtail regulatory risk, empower agents to resolve issues and improve customer experience, and reduce costs is now within reach.

‍

‍