How State Trustees used AI to increase agent job satisfaction by 80%

The client:

State Trustees is a crucial institution in Victoria, responsible for managing the financial affairs of individuals unable to do so due to disability or mental illness. Their services include collecting and managing clients’ income, paying bills on behalf of clients, maintaining and overseeing clients’ property and handling legal issues related to clients’ financial matters. State Trustees ensure that vulnerable Victorians receive the financial support and management they need, safeguarding their interests and well-being requiring a highly skilled and resilient workforce.

The challenge:

In 2020, State Trustees faced a significant challenge within their call centre, exacerbated by the Covid-19 pandemic. The nature of the calls received were highly stressful, as clients often urgently sought access to their funds. And when funds were unavailable, callers frequently became extremely abusive which created a toxic environment for call centre staff. The Covid-19 pandemic heightened emotions, intensifying stress in an already demanding support service leading to high rates of staff absence and attrition, primarily due to mental health issues arising from the abuse.

State Trustees were facing a legal Occupational Health and Safety (OHS) problem.

The key issues:

  • High levels of abusive language directed at staff.
  • Staff burnout and mental health concerns.
  • Increased turnover and absenteeism.
  • Difficulty in managing escalating calls effectively.

Our solution:

We implemented a multi-faceted solution focusing on automation, staff protection and escalation management.

Automation:

We automated as many processes as possible, diverting routine inquiries to a chat system. This not only reduced the call volume but also provided a buffer for staff from direct verbal abuse. In the chat interface, all abusive words were automatically replaced with ‘#’. This allowed staff to focus on the customer’s query without being

exposed to harmful language. Customers were informed about the omission of abusive language, encouraging more respectful communication.

Analysis:

We employed advanced sentiment analysis tools to monitor the tone and content of interactions in real-time. When a call or chat interaction began to escalate and showed signs of becoming abusive or particularly challenging, it was automatically transferred to more senior and experienced staff. Customers were notified that their interaction was being escalated due to the nature of their communication, which helped to deter attempts at bullying less experienced staff into releasing funds.

The result:

Our innovative approach to reduce absenteeism and attrition focused on understanding and addressing agent concerns to provide a robust model with remarkable results in just 8 weeks.

Outcomes:

  • 6% of all inbound queries successfully deflected to live chat
  • 92% of users expressed their intention to use live chat for future interactions
  • 80% increase in human agent job satisfaction

By understanding the root cause we enhanced the agent experience, leading to better outcomes for both the staff and the clients they served.

Improved Agent Wellbeing

The protective measures led to a notable decrease in staff absence and attrition. Employees felt more supported and less stressed, improving overall morale.

Reduction of Abusive Interactions

The automated chat system and real-time filtering of abusive language significantly reduced the direct exposure of staff to harmful language.

Enhanced CX

The use of sentiment analysis and escalation protocols ensured that difficult interactions were handled by the most capable staff, maintaining service quality and consistency.

Improvement in Customer Behavior

Informing customers about the filtering of abusive language and the escalation process encouraged more respectful interactions, fostering a more positive communication environment.