Actionable Uses for Conversational IVR-generated Data

Get More from Conversational IVR

Conversational IVR is shaping up to be the next major shift in voice communications. It’s easy to understand why. Conversational IVR speeds up customer interactions over the phone, it’s intuitive and simple to use, and helps to deliver positive customer experiences.

This customer-centric view is both appealing and valuable to companies interested in conversational IVR. But the artificial intelligence (AI) technologies that enable conversational IVR generate a lot of data. And forward-thinking companies can tap into that data in order to drive improvements and optimization in some areas that might not immediately come to mind.

Tapping into AI Data

Understanding how conversational IVR works sheds some light on what kind of data businesses can expect. When an end-user calls into a conversational IVR, the IVR app prompts them to take an action. The caller states their objective and the application routes them to the right place. Sounds easy enough, right?

There are a few processes that take place between the time when the caller states their intent and the IVR app routes the call. The natural language processing (NLP) technology powering these apps interprets the caller’s speech and converts what they said to text. This text is then run through an AI algorithm to determine the caller intent. Once the algorithm identifies what it thinks is the correct intent, it then instructs the application where to route the call. This all happens in the span of a few seconds.

Of course, NLP technology isn’t fool proof and it can’t always accurately determine caller intent. Yet, these failures are also useful. Companies can study them to discover where the algorithm struggled and then refine the intent to accommodate these refinements. This is pretty straightforward and what you might expect when dealing with NLP technology.

One might assume, then that companies would focus on the NLP errors and discard the data associated with successful intent processing. But there may be additional value in that data. Those data reflect the way that customers understand what a company (or perhaps an industry) does and how to interact with it.

Linguistic Data

It’s easy for businesspeople to think about customer service from a company perspective, including the language used to engage with their company. But the problem with this is that the more expertise one gains in an industry or with a specific company, the more their specialized their vernacular becomes.

So, to flip this idea on its head, it’s easy for businesses to forget how the average person talks about their company. Or, even worse, they never factored that into the equation in the first place. But the speech-to-text data derived from conversational IVR applications provides a view of the customer’s understanding of the business and the vernacular that they use.

Analyzing this data can help companies develop better customer service and marketing efforts. Instead of forcing customers to talk about a company in the preferred terms of that business, this approach puts the voice of the customer front and center and builds out from that. In this context, voice of the customer quite literally refers to the word choices and combinations that customers use. Incorporating these linguistic tendencies into your IVR prompts, the scripts for your agents, and marketing copy and messaging engages customers in a more familiar way.

This type of process may also provide another means for companies to expand their AI portfolio because conversational IVR can produce a lot of data. Effectively analyzing that data might require another AI workflow. IVR has always been a gateway technology because it’s foundational to so many communications channels. This trend continues to be true as we shift into the realm of conversational IVR. In the era of big data, the companies that are able to effectively use the data they have will be at an advantage. IVR can be a valuable component of that process.