The BPO Revolution: From Agent-Assist to AI-Assist
Introduction
Traditionally, agent-assist tools have been used to help call center agents navigate conversations and efficiently handle after call work. However, with AI's rapid advancements, we are now witnessing a shift from "agent assist" to "AI assist," where AI takes on a more prominent role and human agents help guide the process.
Agent Assist Tools: A Brief Overview
A variety of agent assist tools have been developed to enhance agents' performance and improve customer experience. Some examples include:
Knowledge management systems: These tools consolidate and categorize information that agents need to handle customer queries, making it easily accessible during calls.
CRM integration: Customer relationship management (CRM) integration tools help agents access and update customer data and call history in real-time.
Scripting and decision support: These tools guide agents through conversations, providing step-by-step instructions, prompts, or possible responses to ensure a consistent customer experience.
Post-call analytics: Advanced analytics software analyzes calls and generates insights for agent training and performance improvement.
The Shift to AI-Assist
With AI's increasing capabilities, some BPOs are now using AI to handle the majority of tasks in chat, email, and SMS support channels. Human agents are present primarily to provide a "human touch" to the AI-generated responses. This groundbreaking approach has resulted in significant benefits, including:
Eliminating training time: As mentioned earlier, AI's sophisticated understanding of client products and services allows agents to work on client programs without any prior experience or training. This enables even brand-new agents to provide the same or better quality service than provided by trained and experienced agents working without AI.
Dedicated agent quality at shared agent prices: Dedicated agent support refers to having a team of agents exclusively assigned to a single client. In contrast, shared agent support involves agents simultaneously handling multiple clients, which is usually less expensive due to resource sharing. However, shared agent support has traditionally resulted in a lower quality of service. AI tools are now making it possible to use the less expensive shared agent model, while producing the types of high quality results that were previously only seen with dedicated agents.
Flat fee pricing: AI's ability to accurately predict the cost per resolution, based on historical data, allows for flat fee pricing models. This approach provides businesses with predictable and transparent pricing, making it easier to budget and plan for customer support expenses.
Agent revenue sharing: The flat fee pricing model makes it easy to give agents a percentage of revenue from each query successfully resolved. This incentive model aligns agents' interests with those of the company, motivating them to provide exceptional customer service and fostering a sense of ownership and responsibility.
Attrition rate is becoming irrelevance: For traditional BPOs, attrition rates are a major factor considered by our clients seeking an outsource partner. High attrition rates can increase the amount of resources spent on training and keeps average handle times elevated as new agents are continually introduced to the system. But with AI-assist, the risk of attrition is shouldered entirely by the BPO that has committed to the flat rate. The costs of training and increased average handle time simply each away at the BPO’s profit margin. This gives the BPO more incentive to improving the quality of the AI/human agent interaction so that queries are sufficiently. resolved as efficiently as possible with minimum-to-no human agent training.
The State of Play
Right now, we are only seeing AI-assist being accomplished in digital support channels, such as chat, email, and SMS. Implementing AI-assist in voice support, on the other hand, presents a unique set of challenges due to the complexities associated with human speech and real-time interactions.
Unlike digital support channels where there's a slight delay in communication, voice support requires immediate and natural responses. This presents significant engineering challenges that have not yet been overcome. For this reason, voice AI solutions today focus on providing fully automated AI interaction. Some of them will then provide agent assist functions in the event the call is escalated to a human. There currently is no technology that enables an AI-assist business model for voice support.
Let’s Talk
If you would like to learn more about the BPOs leading the AI-support revolution, send an email to jwalter@zmaxinc.com or schedule a 15-minute Zoom call below: