How do you build a business case for speech technology applications?

Building a business case for speech technology applications should be no different than any other technology investment. You need a well-defined use case that addresses a business problem, along with a tangible benefit -- ideally, one that supports ROI. However, speech technology presents inherent challenges that make it more difficult to justify an investment.

The conventional speech technology use case is in contact centers, where speech technology is used for interactive voice response (IVR). This form of self-service provides several benefits, mostly centered around automation that allows agents to focus on high-value inquiries. And it enables contact centers to handle a higher volume of calls more cost-effectively.

That use case is still relevant in today's AI-driven speech technology applications, though the capabilities now are in the form of chatbots, rather than IVR. Chatbots may be the most widely used application for speech technology, and AI capabilities are a big step forward for contact centers.

Some enterprise use cases are emerging that fall under the unified communications (UC) umbrella. AI-driven speech technology applications, like speech-to-text and text-to-speech, can be evaluated based on how productive they make workers. The idea is that AI opens up a rich set of use cases that provide new forms of business value. The first step is to identify what those new values are, as learning curves are expected. But if the organization has already embraced UC, then building a business case will be much easier.

It's important to note that speech technology applications are complex and evolve quickly, which should be reflected in your business case. Speech technology is not yet mature enough to act as a full proxy for human speech, but it's certainly capable of simple tasks. The level of business value may have limits, but AI technologies steadily improve over time. The ROI angle here needs to be different than traditional metrics based on mature, legacy technology that has a longer life cycle.

Do you have a question for Jon Arnold or any other experts? Ask your enterprise-specific questions today! (All questions are treated anonymously.)