BPA Featured Article

Speech analytics can go a long way toward driving improvements in contact center environments. But that’s not all.

This technology can help businesses understand what needs to be improved in terms of agent behaviors, company processes, and product performance. Businesses can then use that information as they coach customer-facing workers, fine-tune how they do things, and upgrade their solutions.

Speech analytics also can help organizations spot customer frustration. That can help them prevent customer churn buy getting a more experienced agent on the line, contacting the individual with a special offer or apology, or some other action.

Creating new revenue is another potential benefit of speech analytics. When certain words or phrases are spoken, they can signal a customer is a ripe prospect for a specific offer. That can create new cross-sell and up-sell opportunities. And catering to customers with the right solutions at the right time both builds walletshare and customer loyalty.

But while we talk a lot about the potential for speech analytics in the contact center and the use of this technology to address customer service situations in real time, Forrester (News - Alert) research released this year says the use cases for speech analytics are expanding, and real-time capabilities are just getting started.

“While many of the vendors in our evaluation offer their solutions primarily geared toward a contact center use case, several of our top-scoring vendors also demonstrated product improvement, marketing optimization, and sales optimization as use cases for their solutions,” said Forrester. “With more vendors specializing in non-contact-center use cases, we’re likely to see an expansion of potential use cases for this technology as the market matures.”

Aspect, CallMiner, Clarabridge, Cogito, Genesys (News - Alert), Invoca, Mattersight, NICE, OpenText, Tethr, and Verint are the most significant AI-fueled speech analytics solutions providers today, Forrester said. (NICE inContact and Verint (News - Alert) are among the companies represented at the upcoming The Future of Work Expo, which will focus on how artificial intelligence is challenging and enhancing business communications, customer service, HR, marketing, and sales.) The research firm added that more than two-thirds of the vendors it evaluated offer some real-time functionality within their solutions, but the implementation of such solutions “varies widely,” Forrester said.

The company added that another key takeaway of its 2018 speech analytics research is that transcription models need tuning. “We conducted an audio transcription test as a key input into the language models criterion and discovered that many out-of-the-box transcription models easily make mundane errors (‘let me get’ mistakenly transcribed to ‘let me guess’) and sometimes run wildly off course (‘microelectromechanical’ turned into ‘my girl went from the town hall’),” Forrester explained. “Additionally, while alphanumeric capture has been bulletproof in the phone channel for IVR, when it comes to speech analytics, the functionality is still catching up.”

Verint says the keys to a successful speech analytics program include:

• Identifying the data critical to your business, and where this data is stored;

• Educating your organization about how speech analytics works, how accurate it is, its features, how to implement it, and its near-term benefits;

• Agreeing on a framework for category building;

• Setting up communication processes so employees know where and how to ask questions, input requests, and offer speech analytics-related suggestions; and

• Acting on the insights

Opus Research (News - Alert) says that 72 percent of companies think speech analytics can allow for improved customer experience. DMG Consulting suggests companies can see speech analytics ROI in one year.