How to Optimize Self-Service Options to Create Better AI Assistance: IVR Best Practices, from DTMF to Conversational AI
Contact centre triaging and routing is critical to understanding customer needs and matching them with the most qualified human or virtual assistance.
Yet, IVR’s reputation is generally negative. From the frustration of a complex DTMF menu to a natural language interface that refuses to recognize what you say. Will customers remember their service interaction as fast and frictionless or a journey of aggravation and repetition?
Inefficient and complex IVR systems negatively impact customer experience and cost contact centres both time and money
IVR services can be a black box experience for customers and organizations to identify those invisible causes of customer frustration that can negatively affect CX.
Common issues include downtime, customers trapped in an IVR loop, slow system response times including ‘dead air’ and prolonged authentication times due to lagging CRM lookups.
Of course, the customer experience goal is to ensure that every IVR system works as intended, which means your self-service options should deliver expected outcomes, but this isn’t always the case.
As a result, IVR systems have gained a negative reputation in customers’ minds reflecting the many challenges of making this service experience happen according to plan.
So, what can go wrong? Take capacity issues for example; when calls can’t get through, or customers must wait either before accessing the IVR or afterwards due to insufficient agent availability, calls are then disconnected.
Maybe your cloud-based IVR is not as dynamically scalable as you thought, or latency interferes with the delivery of voice prompts after DTMF input.
In natural language-driven IVRs, insufficient processing power can impact speech recognition effectiveness and incorrectly predict user intent.
Equally, the same outcome might result from poor audio quality. Other instances include when a DTMF dial pad stops working and becomes nonresponsive to user input. Therefore, the IVR fails to register user input and repeats a prompt instead of advancing to the next one in the user journey.
In all these instances, automated end-to-end testing delivers the actionable insight needed to fix things. At the same time, ongoing monitoring ensures that problems get fixed fast.
Co-Ordinate Your Approach to IVR Design and Testing
Being able to offer automated testing at scale enables the comparison of performance differences between periods of low and peak demand. Points of failure and root causes are identified and checked against originally recorded calls, while ongoing monitoring transforms your responsiveness to issues.
Take a case study in banking, for example. Heavily personalized IVRs can lead to complex test cases. For instance, callers can only listen to the menu related to their account because each financial product has its own self-serve.
On top of that, multiple customer segments will have different routing priorities, impacting the customers’ connection speed to an agent.
Can payments be made to collections lines? If those lines are down, customer credit ratings could adversely be affected, leading to litigation if proved that access was not available. The FSA has some pretty stiff penalties around duty of care not being met.
The latest best practice for IVRs combines functional reliability and dialogue effectiveness. As a result, teamwork focused on these complementary skills is now a priority because organizations must evolve their triaging and routing to align with fast-changing customer needs.
Such as the rise in digital lifestyles, the surge in demand caused by external factors and the emerging cost of living crisis. Customer vulnerability is, therefore, now a leadership priority.
To help you navigate how to best design and test your IVR system Hammer hosted a live roundtable with industry experts.
Martin Hill-Wilson moderated an open discussion that brought together the latest best practices in contact centre testing: Mark Kowal, Product Marketing Director at Hammer and conversational design with Hans Van Dam, CEO at Conversation Design Institute.
We explored how these skillsets can start working together with the common objective of enabling fast and frictionless access to customer outcomes.