ICMI is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.


The Logistics of Implementing Speech Analytics for Your Contact Center

Much of the documentation about speech analytics involves the capabilities of the solutions and a largely theoretical analysis of how it can benefit your contact center. However, implementation of these systems often requires a more detailed approach. The logistics of choosing and installing this powerful new technology can make a substantial difference in its effectiveness.

Evaluating Your Contact Center’s Needs

To successfully implement speech analytics, you must first carefully examine your contact center’s needs. Though speech analytics was once limited to larger contact centers, and primarily focused on automatic categorization of customer interactions, a logistical plan must go beyond that elementary capability.

Contact center management, supervisors and agents must collaborate about the potential use of speech analytics in three distinct areas: classifying customer demands, investigations and opportunities, and monitoring contact center traffic.

Evaluating Speech Analytics Uses

Demand analysis helps to determine what your customers are requesting. It can spot gaps in contact center service or detect a shift in the issues discussed during customer interactions. For example, a lot of people asking for assistance with “password resets” could indicate problems with the same service on the company website. Similarly, if the number of unclassified calls grows, a new product or campaign may be creating confusion.

Investigations flow from the automatic categorization and analysis of customer requests. They may suggest a change in processes, deficiencies in agent training or more serious product flaws. They can also suggest opportunities, such as adding new features or capabilities to your service.

In addition, using speech analytics to monitor contact center traffic may yield simple and effective results. The ability to automatically categorize and count different types of calls can provide significant returns on investment by pinpointing the most critical operational needs.

Evaluating the Technology

Once you have determined the goal(s) for your speech analytics solution, you need to take a closer look at various products. Speech analytics products can be categorized into keyword spotting (KWS), phonetic indexing and transcription-based analytics (TBA).

KWS, the simplest and least costly, is limited to searching for pre-defined phrases while phonetic indexing lets the user search for any word by using a phonetic index; both approaches are phonetic-based approaches (PBA). TBA converts the words into text, allowing integration with other company systems.

PBA provides the most flexibility for monitoring customer interactions and quickly detecting events such as a spike in calls. TBA is typically used for more complex and time consuming root-cause analytics.

Evaluating Personnel and Process

After evaluating your goals and the various technologies, you must determine if your staff can handle the system. In order to be useful, speech analytics must be properly applied. Someone must be able to use it, and someone must act on the information it supplies.

You must also design a process to meet your contact center’s needs as described above. Ask and answer basic questions such as, “What are we trying to achieve? Who does what? What will we do with the information?”

A carefully thought-out and strategized plan will yield the most useful and voluminous results

The logistics of implementing speech analytics involves a step-by-step approach to evaluate your contact center’s needs, personnel, technology and processes. Contact centers following this formula will benefit the most from the implementation of speech analytics.

Marco Mueller is the Executive Vice President, ASC telecom AG.