Skip to content

For Speech Analytics, the Best Is Yet to Come

For Speech Analytics, the Best Is Yet to Come A top-10 list for this transformational solution.

8/12/2013
By Donna Fluss
Speech Technology

  Printer Friendly Format       View this document on the publisher’s website.

In the nearly 10 years since speech analytics has entered the technology marketplace, it has become an essential transformational solution—not just for contact centers, but for the entire enterprise. It is compelling for several reasons:

  1. It addresses a real challenge—gaining insights into customer needs and wants.
  2. It delivers quantifiable benefits.
  3. There is nothing else like it in the market.
  4. It improves the contributions and payback of other solutions used in most contact centers.

Now that organizations around the world have figured out how to use it, the best is yet to come.

Top 10 Applications

Contact centers are the primary users of speech analytics, even though these solutions make valuable contributions to other operating departments as well. The number-one use for speech analytics in contact centers is identifying the reasons people call. These findings are not actionable until combined with other metrics, such as first contact resolution (FCR).

The second is automating the QA process. Increasingly, vendors are selling analytics-enabled QA solutions in which speech analytics identifies calls for review, instead of the traditional manual approach.

As contact centers remain productivity-oriented, even in the customer experience era, improving FCR rates by decreasing the number of calls that are transferred, placed on hold, or require a follow-up callback is the third most common use. A pure productivity play, and the fourth most common use of speech analytics, is reducing average handle time. This is surprisingly easy to achieve. Managers can use speech analytics to identify calls that are longer than most. These can be filtered by criteria such as call reason and emotion level. Managers can concentrate on call subsets to identify underlying reasons for long calls and address those issues.

Script compliance is the fifth most common use. This is where companies use the solution to ensure that agents have communicated necessary information to a caller or that an agent or collector did not say prohibited things. (A proliferation of regulations is making speech analytics a must-have for collections organizations.)

It is common to use speech analytics to identify training opportunities for agents, the sixth most common use. The true value of this application lies in the ability to spot specific issues an agent is having, whether it’s keeping talk time down or handling challenging customers.

Using speech analytics to improve the collections process is the seventh most common use. Collectors must comply with strict requirements, including identifying themselves, reading a mini-Miranda, telling the person they reach that the call is being recorded, and confirming that they are speaking to the right person, all within seconds of being connected. This is hard to do, but can be learned once the issues are identified with speech analytics.

The eighth most common use is increasing sales. The potential for this application is high for organizations willing to rethink how they do business. Speech analytics can complement focus groups or uncover new ideas from customer feedback. It can also help organizations prioritize what they sell and how they sell it. Even better, real-time speech analytics can kick off a guidance work flow that gives agents the information they need to close more upsell and cross-sell opportunities.

Another application of speech analytics is identifying problems with self-service solutions, such as a voice response solution or Web self-service environment. Customers share their thoughts openly about these environments, so this is an ideal use for speech analytics. There are many ways organizations can improve self-service applications once they are aware of the issues.

Reducing customer attrition is the tenth most common use of speech analytics today. This application can produce a rapid payback, but the process can be complicated. While speech analytics can identify at-risk callers in real time, not all customers should be saved, and organizations should have clear procedures in place so agents know what actions to take. Additionally, speech analytics can analyze historical data to determine proactively when customers are at risk of closing an account. Once they are identified, the organization can engage in a campaign to retain them.