Many companies and organizations that already use text-based chatbots are now embarking on a quest to implement voice-bots by adding voice and telephony channels to these existing chatbots. But the actual integration of the voice and telephony channel, which often poses technical difficulties in itself, is just one part of the equation. There is another equally important aspect that organizations will certainly come up against during the implementation process – adjusting the user experience.
Facing the UX Challenges Head-On
The overall user experience of the voice-bot is influenced by the bot’s capabilities, and the optional interaction the solution enables with the telephony world. There are three main challenges affecting the UX when connecting bots to telephony:
- While the telephony call is happening, many events are taking place and telephony data is accumulating. For a voice-bot to deliver a good user experience, these events and data – such as the calling number, caller name, customer data gathered during previous interactions, dialed digits (DTMF) and more – must be integrated smoothly into the call. All this can change the voice-bot user experience dramatically.
For example, when a contact center receives a call, a correctly deployed voice-bot can identify the customer’s number, retrieve his or her name from the database, analyze previous interactions and then embed all this information into the call. In such a case, the bot might answer the call by saying: “Hello Mrs. Lee, how can I help you?” It can then refer to previous interactions later on in the call if necessary. You can see a good demonstration of how this works in our video.
- The voice-bot initiates activities as well. It does not just react to data that was gathered in the past, but also contributes relevant metadata that was captured during calls to the contact center. Other bot-initiated activities that are relevant to telephony calls include things like call recording or implementing additional voice use cases, such as passing the call on to a human agent, when required. In this case, after notifying the customer, the voice-bot will redirect the call to a human agent, along with all the relevant data that was gathered and is relevant for the call.
- Inputs that are not relevant to voice – such as visual menus, emojis and images – should be eliminated. In addition, the bot needs to learn to ignore situations that do not happen in chat interaction, such as caller interruptions and no user input.
If these three challenges are addressed properly, you will significantly improve the chances that your voice-bot will delight your customers with an intuitive user interaction, all while slashing your operational costs.
Here’s the Easy Answer!
We at AudioCodes developed VoiceAI Connect to create an effortless bridge between any bot framework and any telephony system to support all major voice-bot use cases. Hundreds of bot developers and companies are already using this innovative solution and dozens of leading partners in this domain are participating in our eco-system and community.
So, if you want to implement a voice-bot and use the knowledge and expertise accumulated in the VoiceAI Connect eco-system to beat the challenges described above, you’re welcome to get on board and join us.