WHITE PAPER
Speech Analytics for Identifying Agent
Skill Gaps and Trainings
©Uniphore Software Systems | Contact: [email protected] | Website: www.uniphore.com
1
Table of Contents
Introduction .................................................................................................................................................. 3
Problems ....................................................................................................................................................... 3
Speech Analytics - solution .......................................................................................................................... 4
Effectiveness of speech analytics............................................................................................................. 6
Conclusion .................................................................................................................................................... 9
©Uniphore Software Systems | Contact: [email protected] | Website: www.uniphore.com
2
Introduction
A challenge for any contact center is to understand the agent skill-gaps and guide the agent through
training based on their skill set. In contact center industry, speech analytics is one of the fastest growing
applications to provide a complete picture of the agent’s performance on real-time, helping them
perform more efficiently and giving them real time guidance. Speech analytics has the potential to
positively impact contact center agents across industries and provide a way for the contact center to
analyze and measure the effectiveness of your on-boarding and contact center training programs, and
track the progression of agents through their careers.
Problems
As data analytics tools have become more advanced and ubiquitous, pinpointing an agent’s skill gap can
be complex. Identifying the agent skills through careful assessment, mapping a plan to fill the gap,
arranging the training based on their skill sets and identifying measurable outcomes are very cost
effective and time consuming. Traditionally, performance information of each agent is stored in
different systems and correlating data points is time-consuming or difficult to achieve. It is also very
complex to keep tracking of the actual skills of agents in contact centers of all sizes operating in different
business domains.
In terms of skills shortages and gaps in contact center, the agent may lack in customer handling skills,
soft skills, technical and practical skills, team working skills, oral communication skills and problem
solving skills. All of these skills reflect essence of working in a contact center environment. The agent
should be strong in soft skill when dealing with customers either face-to-face or via other
communication methods. In terms of technical and practical skills, these are necessary to ensure that
the methods and processes of contact centers are in place in order to give customers a good experience
leading to customer satisfaction and loyalty.
The use of this traditional data analytics tool contact center could only perform post-call evaluation.
Pinpointing a skill gap where an agent could improve was inexact and time consuming and many critical
situations simply went unnoticed. It is very difficult to evaluate each and every call handled by the
agents, producing inaccurate results and frustrating both agents and supervisor. For example, five calls
evaluated per agent per month is quite simply not enough analysis to get the job done. Since these calls
are randomly selected there is no consistency in the types of calls monitored and the subsequent
insights earned from analyzing these interactions may not be an accurate representation of how that
©Uniphore Software Systems | Contact: [email protected] | Website: www.uniphore.com
3
agent really acts during a call. A review of these few calls would indicate the agent is under-performing,
but in reality the agent could be a top performer. With a result of these calls, managers can tailor their
coaching sessions to handle the agent’s actual needs and issues, which may or may not be an accurate
representation of their real weaknesses.
Speech Analytics - solution
Speech analytics software enables contact centers drive critical business goals by analyzing existing
speech data to build statistically strong customer response function models and enrich it with ongoing
data to predict outcomes and prescribe likely successful customer engagement process. It aims to offer
integrated mining and analytics solutions for spoken conversations to provide insights for improving
agent quality and customer experience, identifying revenue generation opportunities and managing
risks and compliance.
Speech analytics software helps the contact center to identify and understand the agent behaviors and
attitudes in real time based on keywords, call flow and emotional parameters on both the customer and
agent side of an interaction. This information can then be leveraged to create an effective training plan
and improve the agent skill sets in order to meet the critical workforce optimization challenge.
Technology involved is keyword spotting engine and proprietary recommendation algorithms which
helps identifying and prescribing improvements for each business goal.
Through the use of speech analytics, contact centers can automatically identify areas in which agents
may need additional training or coaching, and can automatically monitor the customer service provided
on calls. With speech analytics 100% of agent calls are automatically scored, providing objective and
consistent feedback on every call. This way, agents get fair and consistent evaluation into where they
are doing well and where they need to improve.
With a speech analytics system, contact center agents have the potential to become far more efficient
when working with customers. It helps the agents to get immediate feedback on each and every call
with a customer to improve their performance and meet the customer expectation. Real time
monitoring solutions built on speech analytics technology provides immediate, continuous feedback to
managers, supervisors, and agents during the live call. This means that agent’s behaviors can be
corrected when they are still on the call with a customer, providing next course of actions to take.
Speech analytics can also alert mangers to an unfolding situation, like the customer threat – legal on the
call, before the call ends so they have the chance to resolve the issues immediately.
©Uniphore Software Systems | Contact: [email protected] | Website: www.uniphore.com
4
Below are the key areas where the contact center agents, and the center as a whole, can reap the
benefits of speech analytics.
Real time identification of agent behavior and improve business process
Speech analytics analyzes agent’s conversation with the customer accurately in real time to learn pain
points along a customer’s journey and provide proactive coaching to the agents to ensure that an issue
is resolved to the customer expectations. As companies learn more about the issues that they need to
deal with, agents are provided with the information they need in order to manage calls as efficiently and
effectively as possible, resulting in increased agent performance and happier customers overall. In
addition, the agents work through the suggested real time alerts, learning how to handle this type of call
in the future, potentially saving the company millions in fines. This helps to retain more customers, build
strong customer relationships, and improve overall contact center’s business process.
Accurate Evaluation of agents skill-gaps and arrange training courses/programs
Speech analytics provides a mechanism for contact centers to quickly review an agent’s entire call long,
find the agent’s skill gaps in which they need most improvement, and build a training plan. This means
that a complete picture of the agent’s performance on all of their calls can be produced. However, an
integrated agent dashboard can allow an agent to view the score of each performance parameter and
trends of each parameter for improving the agent’s behavior. It gives a chance to the agents monitor
their own work, compare their own performance to their goals, take the required steps needed to
progress and maintain their own performance. Using these same analysis techniques, contact centers
can also identify and quantify the best practices of the top performing agents and turn those actions
into metrics for others to use.
The training courses/programs reports are created immediately based on the agent’s performance on all
of their calls to change the agent’s behavior for the better and create the best team, resulting in contact
centers can create more consistent and positive agent behavior across the board. Through these
accurate and timely agent evaluations, the agents can be trained across the industries in order to
manage and improve the customer experience, resulting in impact of companies top line revenue and
bottom line profit.
©Uniphore Software Systems | Contact: [email protected] | Website: www.uniphore.com
5
Effectiveness of speech analytics
The below charts describing the agent classification and training requirement across the industry based
on their skill sets.
Training requirements reports through speech analytics
Report: 1
©Uniphore Software Systems | Contact: [email protected] | Website: www.uniphore.com
6
Report: 2
Skills
Agent Classification
Agent A
Soft skills
Agent B
Agent C
High
Medium
Low
Level of probing
Pitching skills
Customer handling skills
Team working skills
Communication skills
Problem solving skills
©Uniphore Software Systems | Contact: [email protected] | Website: www.uniphore.com
7
Report: 3
Contact centers of all sizes can leverage the power of speech analytics software to gain valuable insights
into how business processes affect the agent training.

Recommendation engine is a salient feature of the speech analytics product. It recommends a
pop-up window displaying valuable tips, guidance or assistance to the agent on the contact
center software screen during a live call. This is a way that the contact center improves the
agent’s performance and maximizes the revenue generating opportunities.

The color of the pop-up window will toggle from green to amber to red depending upon the
severity of the alerts and their types. For example, if a normal going call where everything is
going fine, the color of the pop-up window may appear green. However, for calls which have
escalation or legal language from the customer may turn from green to red. For minor alerts,
the color of the window may change to amber. It enables the agent to quickly check for the
severity of alerts and make aware of the proceedings of the call. This is a better way for the
agent to realize that the call needs course correction depending upon the alert, resulting in cost
saving and better customer experience.
©Uniphore Software Systems | Contact: [email protected] | Website: www.uniphore.com
8

Speech analytics solution allows the agent to check his performance parameters like average
quality score, critical errors, customer satisfaction, sales target and so on. Along with the
parameters scores or values, the agent can also monitor the trends for each performance
parameter for the selected time period. The trends can be improving, weakening or stable when
compared with the time period as selected by the agent.
The core technology used in the speech analytics product is Automatic Speech Recognition (ASR) –
keyword spotting engine. Smart event rules are formed for the purpose of scoring, recommendations,
MIS, reporting and triggering alerts based on the keyword spotting technology. These event rules can be
tuned based on the requirements of the contact center to generate higher revenue opportunities and
build strong customer relationships for their contact center.
The simplest and most cost-effective route to well-defined journeys leverages speech analytics as the
foundation. Rather than working to custom integrate multiple vendors or channel solutions, leveraging a
unified speech analytics platform that identifies agent skill gaps in real time and delivers assistance or
guidance during the live call to improve the agent performance and meet great customer experience.
Conclusion
Speech Analytics is a powerful software that has the potential to change the agent behavior effectively
and provide the real time guidance or assistance in order to create more efficient agents and happier
customers overall. It helps the contact centers to get a better understanding of how their agents truly
act during interactions with customers, as well as provide accurate feedback and customized training
solutions to address the weakness of each individual agent. Through the power of speech analytics,
contact center agents identify market intelligence to improve their product or service offerings and drive
business with higher revenues and lower costs.
©Uniphore Software Systems | Contact: [email protected] | Website: www.uniphore.com
9