Speech Analytics: Mining Business Insight from Customer Voices


Share on LinkedIn

Speech analytics can help enterprises listen to and act on the insight stored in the actual voices of customers. In big companies, finding and fixing a problem early can improve customer satisfaction and loyalty or save millions of dollars by eliminating unnecessary phone calls.

But there’s a huge opportunity in smaller businesses, too. Speech analytics solutions are becoming easier to implement and use, and more cost-effective over all.

This year’s SpeechTek conference in NYC gave me the opportunity to catch up with some of the leading vendors in the speech analytics space. In a nutshell, what I learned was that vendors are working hard to 1) apply their technology to business issues outside of the contact center, 2) package solutions designed for business users, and 3) address implementation/adoption issues to penetrate the largely un-tapped SMB market.

Trend No. 1 — Not Just for the Call Center Anymore

Historically, calls have been recorded to help optimize agent performance. Companies listen to call recordings to identify opportunities for agent coaching—to improve productivity, customer satisfaction, compliance, etc. As a result, some vendors place speech analytics as part of so-called “quality monitoring” (QM) or “workforce optimization” (WFO) solutions.

That’s a limiting position. What if a customer calls because of a problem with some other part of the organization? Business success is not just about optimizing the call center.

Jeff Schlueter, VP of Marketing at Nexidia, told me the story of a large multi-media telco that used speech analytics to identify a spike in repeat calls that occurred 30 days after service activation. Turns out customers couldn’t find a promised refund because it was buried on page four of their bill. Moving that crucial bit of information to the top page saved the telco $4 million per year, says Schlueter.

Optimizing self-service usage is another application. Customers that can’t find what they need on the web site, have problems with an e-commerce checkout process or bail out of the IVR are common examples. Phone calls to agents will provide clues to what’s driving these calls, so the root cause can be addressed.

Of course, discovering doesn’t mean fixing. Too often big companies are still dominated by organizational silos and turf wars. But some speech analytics tools can help find and route problems to the right department for handling.

Trend No. 2 — Packaged Business Solutions

Speech analytics technology has earned a well-deserved reputation for being hard to implement (shades of the early days of CRM). Techies are required to install, set up categorization schemes and so forth.

The technology “under the hood” is complicated stuff, to be sure. But business users aren’t particularly interested in whether a solution uses phonetic, full transcription, or some other technology to get the job done.

NICE Systems, for example, now offers packaged solutions that focus on key business issues. First-call resolution (FCR) is one, because it’s critical to call center performance. Churn is another. According to NICE’s Corporate VP/GM Barak Eilam, the UK-based telecom Orange used speech analytics to identify consumers likely to churn so they could act before the customer left. He says that 75% of the churn “alerts” were not discovered by other analytic approaches. And that the source information was not just phone calls, but also included chat and email content processed via a text miner.

NICE Systems is in a, well, nice position to deploy a combination of speech analytics technologies along with text mining (which uses different algorithms). And that’s not all—their predictive churn models were developed with a combination of SAS analytic tools and their own code.

But still, what’s important here is that the technologies are being packaged to address a business issue. And increasingly, vendors are also offering solutions that can be used immediately, sometimes via a hosted offering. That should help increase penetration outside the call center and with smaller businesses.

Which leads me to…

Trend No. 3 — Entering the SMB Market

The low-hanging fruit of the QM/WFO market has been big enterprises. But this market is becoming saturated and commoditized. So where is future growth going to come from?

Smaller businesses could reap the rewards of speech analytics, but solutions have to be much easier to implement and less costly. For a large enterprise, speech analytics can add hefty premium of 50-100% over basic call recording. And that’s just for the software.

Verint Systems entered the WFO market in 2007 by way of a merger with Witness Systems. Like other vendors, Verint isn’t shy about pushing its technology innovation. But Diego Lomanto, Senior Solutions Marketing Manager, agreed with me that SMBs don’t care about what’s “under the hood,” they just want to solve a business problem quickly and cost effectively.

To that end, a couple of months ago Verint introduced an “Essentials” solution with new features designed to help business users use immediately to spot trends and a “guided search” capability for call recordings. Lomanto says Verint’s solution uses the same core technology but is packaged and priced for smaller businesses. And, to help on the cost issues, speech analytics is only a 25-35% adder to basic call recording solutions, he says.

Adoption Challenges

These trends are encouraging, but for broader acceptance speech technology vendors need to do a better job of “getting the word out” to a business audience. I suspect that most business managers would assume anything called “speech analytics” should be delegated to the call center manager or CIO.

Another issue that will prevent reaping the full reward from this technology is organization silos and conflicting goals/measurements. For example, the marketing department may be measured on lead gen and thus not terribly motivated to fix web site design problems that cause agent phones to ring. A senior manager with a multi-function view of the organization needs to take overall ownership and set priorities.

This trend towards business applications and smaller business could raise competitive tensions with enterprise software companies that have traditionally focused on BI, EFM and cross-channel experience management. We’re likely to see the traditional CRM and call center industry landscape reshaped over the next couple of years as vendors jockey for position.

But in the end, I believe this will result in more usable and cost-effective speech analytics solutions that will be more widely adopted across the enterprise, and at smaller companies.

Further reading:
* Envision vs. Verint “Smackdown” on Speech Analytics for SMBs, but the Cloud is up Next
* Use Speech Analytics to Reduce Calls That Frustrate Customers and Hurt Productivity


Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here