BankingConversation Intelligence

Case study

Increasing QA oversight while reducing review costs

A major UK retail and commercial bank used Conversation Intelligence to extend QA oversight from under 5% to 100% of mortgage conversations while cutting review time.

Major Retail and Commercial Bank — UK

London skyline through high-rise office windows at sunset
  • From less than 5% to 100%Mortgage sales oversight
  • Over 50%QA review time reduction
  • 95%Human-AI agreement rate
  • 5,000Conversations per month

Summary

Key outcomes at a glance

  • Oversight of mortgage sales increased from less than 5% to 100%
  • QA review time reduced by more than 50%, reducing compliance costs and generating longer-term savings
  • Advice standards improved and complaints reduced
  • Face-to-face, telephone and video banking conversations could be analysed as part of the bank's mortgage QA process
  • Risk scoring helped identify the cases most in need of human review

The challenge

A major UK retail and commercial bank was conducting approximately 5,000 mortgage conversations every month. While its telephony provider already recorded conversations, the bank needed a way to capture face-to-face client-adviser conversations and analyse them at scale alongside its existing recordings. 

This created a gap in oversight for face-to-face discussions. The bank wanted greater visibility across its mortgage conversations so it could proactively identify trends and inherent risks at the frontline of the business and mitigate the risk of consumer harm before it occurred. It also needed to make the Quality Assurance (QA) process more efficient, enabling greater oversight without creating an unsustainable review burden. 

Advice conversation oversight increased from less than 5% to 100%, while QA review time fell by more than 50%

The Recordsure solution

Recordsure technology enabled the bank to capture, analyse and review mortgage conversations across face-to-face meetings, telephony and video banking. The solution centred on Conversation Intelligence, bringing different conversation channels into the QA process and using AI-driven analysis to help identify conversations requiring closer human review. 

Bringing mortgage conversations together

Recordsure’s Conversation Intelligence enabled advisers to securely record face-to-face mortgage meetings. The recordings were then uploaded and analysed alongside conversations from the bank’s existing telephony and video banking recordings. This gave the bank oversight across its different mortgage conversation channels and addressed the previous gap around face-to-face interactions. 

Prioritising conversations for human review

Mortgage conversations were risk-scored to help identify the cases most in need of human review. The risk scoring achieved a 95% user-verified human-AI agreement rate, helping the bank direct reviewer attention to cases requiring closer examination. 

Making QA reviews more efficient

AI-driven analysis segmented mortgage conversations into colour-coded QA review checklist topics. QA teams could then use these topics to work through conversations more methodically and efficiently. This helped QA reviews be completed in half the previous time. 

The outcomes

Mortgage sales oversight increased from less than 5% to 100%, giving the bank greater visibility across its mortgage conversations. At the same time, QA review time fell by more than 50%, helping the bank increase oversight while making the review process more efficient. 

The bank also improved advice standards and reduced the number of complaints. QA and Compliance costs were reduced, generating substantial longer-term savings. Risk scoring helped identify cases most in need of human review, while structured conversation analysis enabled QA teams to work through reviews more methodically and efficiently. 

Engagement snapshot

  • Since 2018With Recordsure
  • 5,000Total conversations per month
  • Approx. 60,000Total conversations per year
  • 411Users

Frequently asked questions

Recordsure technology enabled the bank to capture face-to-face mortgage meetings and analyse them alongside existing telephony and video banking recordings. This increased oversight of mortgage sales from less than 5% to 100%. 

QA review time reduced by more than 50%. Structured conversation analysis helped QA teams work through reviews more methodically and efficiently. 

Recordsure risk-scored mortgage conversations to highlight cases most in need of human review. The risk scoring achieved a reported user-verified human-AI agreement rate of 95%. 

Face-to-face meetings were securely recorded and analysed alongside conversations from the bank’s existing telephony and video banking recordings, allowing the bank to review mortgage advice conversations across its communication methods. 

The bank reported improved advice standards, fewer complaints and reduced QA and compliance costs, alongside substantial longer-term savings 

Download this case study

PDF · Free

Download Increase QA oversight and reduce QA costs – Major Retail and Commercial Bank – UK (PDF, 325.5 KB)