High customer satisfaction, lower costs: The untapped potential of voicebots in customer service

“Voicebots are software systems that communicate with end customers in natural language,” explains Alexander Witte, Head of Digital Business Unit at Majorel Germany and Eastern Europe. “For example, they can query information, classify concerns, conduct independent customer dialogs, and perform assistance tasks of various kinds.” When using a classic IVR system, the end customer has to navigate through various menu options using a dial key or voice and is then forwarded to a customer service representative on a topic-specific basis. “A classic IVR makes limited use of today’s digitization potential, because it is limited to pre-programmed options and phrases and can be very cumbersome, lengthy and inefficient for the end customer. The use of voicebots in the telephone channel, not only provides a quick and uncomplicated solution to the customer’s concern, but also optimizes the service experience and thus increases customer satisfaction,” says Witte. By partially automating customer interactions with the help of voicebots, companies also reduce costs in the service center.

Although the integration of voicebots in the telephone channel could greatly improve the customer experience, their potential has hardly been made use of to date. According to an internal study conducted for Majorel by the consulting firm ATKearney, an average of around 27 percent of all customer service activities can be covered by artificial intelligence, such as a voicebot. Another study by software provider NICE found that about 80 percent of all end users are dissatisfied with lengthy menu-based IVR systems. There is also a high willingness to use voicebots, according to a study by the consulting firm Capgemini. More than 50 percent of all end customers have already used a voicebot, for example, to request information or purchase products.

That’s why Majorel joined forces with technology provider Omnibot earlier this year to launch a hybrid service offering that combines the expertise of both companies: OmniBot brings in the technology to implement chat and voicebots, while Majorel adds expertise in resolving customer concerns through human customer service representatives in the service center (Our solution page “Voicebots and chatbots”).

In addition to classifying incoming calls according to various topics, a voicebot can, for example, answer questions based on a predefined FAQ catalog and thus solve simple service concerns directly. In addition, it is also possible to solve more complex concerns for which customer authentication is required. These include, for example, the output of the delivery status of an order or the retrieval of an account balance. The solution thus goes a significant step further than classic IVR systems. A voicebot understands the caller’s natural language and responds accordingly. Whereas an IVR system uses menu options and responds to voice or key input with pre-recorded messages, a voicebot provides natural-language responses to customer inquiries and is constantly learning. Unlike an IVR system, it does not have to preset response options, but can openly query the customer’s request. In addition to Automatic Speech Recognition, i.e. the recognition of pre-programmed expressions, which an IVR system also has, a voicebot works with Natural Language Understanding and Processing. An IVR system can therefore only understand expressions to a limited extent, whereas a voicebot enables the end customer to describe his request flexibly and in his own words. It has high recognition rates even for complex requests. “It is one thing whether a system understands what the customer is telling it. The other is whether it can also process the voice input, provide a suitable answer and thus help the customer directly,” says Witte. The advantage of a hybrid service model of human and technology such as that offered by Majorel and OmniBot: The system forwards complex concerns or those that require human empathy specifically to the appropriate contact person in the service center. An IVR system, on the other hand, is primarily used to prequalify concerns. Here, the customer does not have the opportunity to describe his or her concern directly. It is merely a matter of assigning them to a suitable contact person.

When using voicebots, the service center employee has the option to intervene and take over the call at any time – but only when it is really necessary. The system enables real-time monitoring of the human-machine interaction. Especially at the beginning of the voicebot implementation, in the so-called hypercare phase, customer service representatives regularly intervene. Over time, the voicebot learns and can resolve concerns directly more and more often. In addition, the voicebot is customizable. For example, programming can be set up so that complaints are forwarded directly to a contact person in the service center. The focus here is not only on reducing costs in the service center, but above all on the customer experience. By using voicebots, companies can for example avoid long waiting times. At the same time, in the hybrid service model, complex concerns and valuable contacts still lie with the customer service representative. After all, these should be remembered positively by the customer in the long term.

This article first appeared in TeleTalk 3/2021.