AI works best with highly repeatable, high volume tasks. The majority of our customers see that upwards of 50% of their customer service tickets span the same 5-7 issues. It’s with these issues that AI needs to be used to respond instantaneously, without a customer ever getting into the human agent queue.
When a customer has a complex issue, AI should be an Agent Assistant to prepare information for a human agent to work faster. AI can determine a customer’s intent, clarify or ask for additional information, and pull data in from other systems, before a human rep becomes involved. By using AI to prepare the information, the human agent can focus on reviewing, making a decision and communicating back with the customer quickly. Consumers no longer wait for hours to get a response.
Agents today spend their time doing five primary things:
For the tickets that are routed to a human agent, an AI Agent should assist with the first three steps. Here are a few examples:
- A customer reaches out to ask for a refund outside of the policy-allotted 30-day return window. The AI accesses the CRM to confirm which product the user would like to return and inquire about the reason before packaging all of the information to the human agent for a decision.
- An airline customer is asking for available options on an earlier flight today. The AI pulls up the customer’s loyalty status and confirms he is a Gold Star customer, while confirming his current booking details. The AI simultaneously searches for flight availability, including First or Business class options, and presents the human agent with what’s available for them to decide if change fees should be waived, which upgrades could be presented at no cost, and the best options to offer.
- A CPG customer is reporting an issue with a product. The AI clarifies the issue she is facing and confirms the product SKU before passing off to a human for follow up.
- A hotel customer claims that she has been overcharged. The AI looks up recent history and provides an itemized bill to enable the guest to review. Once the guest confirms that she believes she has been overcharged, the AI confirms the last 4 digits of the credit card and kicks the conversation to a human.
AI can decrease agent turnover
Customer service reps are leaving their jobs at one of the fastest rates of any industry, propelled by stressful, long days in which they often have to deal with less-than-pleasant, emotional customers. When AI is used to respond to the mundane, highly repetitive tickets, the job itself becomes more fulfilling: the focus shifts to high-impact, high-touch work. Furthermore, human agents are empowered with the right information to make more informed, faster decisions. The benefits are clear: Agent work is less tedious; they are more fulfilled and I would predict that this leads to less attrition (where businesses save on the cost savings for hiring and training new reps).
The move to automate customer service does not mean that all support should be managed by AI. The main priority needs to be on providing high-quality resolutions to customer issues quickly and conveniently. Companies should not try to automate issues that AI is not equipped to handle – customers will become frustrated and increasingly annoyed as they try to reach a human on other channels. Automate repeatable customer queries, but also empower human agents with the information they need to more efficiently solve complex and unique issues.
The future is not 100% automation. The future is AI + Human collaboration.