Skip to main content

Using voice and interaction analytics to identify consumers in vulnerable circumstances

Technology is changing the way people access banking and insurance products and services. More than 80% of Australian consumers now prefer to do their banking through electronic methods such as mobile, email, chatbots and online applications.1 Similarly, in the Australian insurance market, limited physical stores mean more consumers are researching, applying and claiming electronically. Applying for a product, making a complaint, or enquiring about a service, can all be done at any time of the day from anywhere in the world. 

Identifying consumers experiencing vulnerability, in a digital environment, is now a very real challenge facing Australian banking and insurance institutions that have historically relied heavily on face-to-face interaction to identify consumers in vulnerable circumstances. 

In this blog we explore how organisations can use the voice and behaviour of the consumer (quite literally) to proactively identify those that may require extra care to effectively meet their banking or insurance needs.

The meaning of the interaction

Observing and analysing interactions with customers is not a new concept, but increasingly over the last decade the application of technology in monitoring these interactions in the banking and insurance sectors has exploded. 

How does it work?

Voice and interaction analytics extract consumer language and behaviour from sources such as call recording, video conference, webchat, social media, and emails. Variable combinations of speech experts, machine learning and AI then review the extracted data to contextualise and analyse words and actions, and provide one-off or continual insights. Key phrases, pauses, ‘overtalks’, coercion, and even sentiment, can be analysed to uncover the true meaning and intent behind consumer actions. Voice and interaction analytics can be deployed in real time or near real time (i.e. once a week). 

By leveraging these technologies an organisation can effectively monitor all customer interactions, which was not possible when relying  only on manual sampling and ad hoc supervision. 

Notably, interaction analytics is not designed to replace frontline staff, but rather be used with other resources to assist organisations  gain a greater understanding of an interaction with a customer (whatever the medium) in the absence of body language. Context is difficult to assess using just technology, especially when vulnerabilities are transient and may result in high rates of false positives. As such, manual intervention is required to assess context, specifically the circumstances surrounding follow-up consumer contact. 

Nevertheless, effective use of interaction analytics can assist organisations to identify consumers who are experiencing vulnerable circumstances so they can provide adequate support and monitor the appropriateness of the consumer outcome. It can also result in improved consumer experience, products and services and financial performance.

What does it look like in real life?

Customer officer discusses the mortgage extension with the customer who explains the money is needed to build a carport.

The customer officer completes the application and submits it to be processed.

The results of the  call monitoring identify that the customer was being continually spoken to from someone off the phone. Due to the presence of another voice, the application was flagged, and the direct banking team were notified that a follow-up call was required to confirm the customer understood the nature and impact of their application.

A member of the direct banking team contacts the customer to discuss their mortgage extension application. The customer requests a call back. The team member contacts the customer at their elected time. During the course of the conversation it becomes clear that the reason the customer gave for the extension was false and in fact, they were borrowing additional funds for a family member who was rejected for a loan.

The application was paused, and the consumer was referred to an appropriate support service that deals with coercive financial control.

How can we help?

Voice and interaction analytics can assist organisations, using consumer voice and behaviour, to identify those in their consumer cohort that may be experiencing vulnerability.

Deloitte has an established Interaction Analytics team who offer voice and interaction analytics as a service. The team can help organisations get the most out of their data, and importantly, assist businesses navigate how to effectively, but compliantly, analyse consumer voice and behaviour to improve their experience. Whole customer journeys can also be added together to provide a holistic understanding of a customer’s experience across the customer’s interaction lifecycle.

Deloitte can also assist you to design and implement processes surrounding interaction analytics and provide actionable insights to support with your decision-making, including understanding who are your vulnerable consumer cohorts and assessing opportunities to enhance staff capability in identifying and supporting different types of vulnerabilities.

Stay tuned for our next blogs where we will explore the establishment of system flags to support vulnerable customers followed by using accessible and inclusive design to protect vulnerable customers in the financial services industry.


1. Australian Banking Association, Banking customers continue shift to digital