As businesses continue to augment contact centre operations with automated virtual agents, new opportunities can be found to improve customer experiences with access to the right data.
When a large Canadian telecommunications company found themselves in this exact position when customers needed to talk to an agent. In many cases, customers end up being transferred from one agent to another, usually due to an incorrect queue assignment from the virtual agent or the need to interact with multiple agents. Not only is this annoying for them, but it puts an extra cost on the business.
The challenge for the company was to reduce the number of transfers, keeping customers happier and saving on call centre costs. Previously, the telecom used a virtual assistant to support its two lines of business (wireless and wireline), with a combined 30 queues covering sales, billing and tech support. The agent-to-agent call transfer rate was too high at around 13.5%.
To solve the problem, the company needed to identify where queue mapping could be more efficient, identify situations where customers were being incorrectly transferred and use an additional step to ensure customers were in the right queue.
Wysdom started working closely with the company to understand all the customers’ steps through queues and study why calls were being transferred so frequently. This approach involved a thorough intent-to-cue mapping exercise to understand customer needs and the call centre responses.
By updating the mapping to reflect how calls were allocated to queues, more calls were placed in the right queue, helping reduce the issue. Also, by adding a confirmation step for users in the agent escalation process, customers could see if they were assigned to the correct queue, with the option to confirm, or decline and transfer themselves to the right queue.
Providing customers with a dash of interactivity and a greater understanding of the process helps to reduce tension, and is always a good thing for a customer. It leads to greater success outcomes and makes the call centre process more efficient.
Within the business, the study also revealed insights into queue escalations and transfers. As a bonus, Wysdom was able to give the telecom company some useful advice about opportunities to improve agents’ skills across specific queues and query instances.
For example, many customers who were in the tech call centre queue were transferred by an agent to the billing queue. On investigation, this was because people had issues with their services and wanted refunds, as a follow-up. The telecom is now working on enabling tech agents to provide low-value refunds without the need to transfer the call. That’s another win for the call centre and the customer!
Through this process, the client found rapid value increases in time, efficiency and value savings. The major impact was a 20% reduction in the internal agent queue transfer rate, with a reduction in incorrect routing and greater efficiency for the call centre.
The mapping exercise based on studying transfer patterns (intent-based routing) helped improve the quality of call centre operations.
The new user queue confirmation step before agent escalation (queue confirmation), improved customer satisfaction and understanding of the system.
Customers with multiple issues to solve that require multiple queues are more likely to have their issues resolved faster.
Any customer service process needs to be regularly monitored, especially as technology takes over. The company and Wysdom plan to assess the further impact of agent cross-training to consolidate natural business process next steps spread over queues. Through continuous monitoring and optimization of queue transfer patterns, we can identify more queue mapping gaps for further action and business improvement.
Looking to supercharge your chatbot?
Most chatbot implementations will fail to meet business needs, unless they are continuously optimized and tuned to your customers’ natural language, intents and KPIs. With our managed chatbot operations, Wysdom can guarantee you that your chatbot will meet or exceed the mutually agreed KPIs.
That’s true Wysdom.