March 1999

Building A High-Performance Call Center Workforce Through A Scientifically-Based Selection System

BY RALPH HAKSTIAN AND LINDA SCRATCHLEY,
HR DECISIONS LTD.

Workforce improvement has become increasingly important to managers in the recent years of swift technological change and intensified competition. Nowhere is this concern more pressing than in today's rapidly changing call centers. The two principal ways to achieve such improvements are through careful personnel selection and on-the-job employee effectiveness training. Training is often where managers focus most of their efforts. Certainly, new-employee training - dealing primarily with systems and procedures - is essential. Subsequent training interventions concentrating on "soft" performance-enhancement skills, however, have had mixed reviews, and the ability of such costly training programs to yield tangible workforce improvement has not been established. If we shift our focus to personnel selection, though, we will see how the organization that uses systematic pre-employment procedures lays a solid foundation for an ultimately high-performing call center workforce, at a fraction of the cost of later training interventions.

Pre-Employment Testing In Perspective
Some call center managers have simply disregarded testing at the personnel selection stage, often as a result of myths which portray testing as inaccurate and somehow dehumanizing. In addition, some managers avoid pre-employment testing because of negative experiences with selection instruments that were too brief and ultimately worthless. The marketplace abounds in examples of what we might term the "20-minute wonder" - the quick and inexpensive general-purpose test. Single measures of this length and quality can do more harm than good because the user, oblivious to the fact that such tests lack the job focus, reliability and relevance to provide measurements of real value, may fail to consider other, more valid, information in the hiring context. By using a systems approach, however, the call center manager has access to far better pre-employment assessment tools.

Over the past four years, we have been conducting research on how to best assess call center applicants and predict success in this line of work. Let's take a look at what has been learned.

1) Look for a selection system that is based on objective, focused research. This is key. Always inquire about the amount and kind of objective research that has been conducted with the test in the focused context of selecting call center employees. Scrutinize the source of the research you review. In-house studies may contain inflated statistics and errors of experimental design that could affect the results. Far better is research published in the scientific literature, where the work must undergo rigorous review by experts before it is published. The best assessment instruments and processes meet this higher standard. Systematic research establishes the critical twin pillars of reliability and validity:

• Reliability - How stable are the measurements? Are you getting a true and lasting picture of applicants, or just how they appear today?
• Validity - How relevant are the measurements? Are you getting data truly related to call center job performance, or merely information that seems reasonable, but, in fact, is unrelated?

Reliability and validity determine the fairness, accuracy, usefulness and legal defensibility of the assessment process. They can be established only by thorough developmental research in a call center context, conducted by the supplier of the assessment procedure long before its implementation in your center.

2) Look for a selection system that is comprehensive. This is where the notion of a system comes in. No single test can provide a truly effective prediction of eventual call center work performance, but a carefully researched set of key measures, properly configured - an assessment system - can do so. We have found that the test battery underpinning such a focused system takes between one and a half to two hours of testing time to measure the relevant dimensions adequately. If this seems lengthy, keep in mind that only those hoping to be hired are investing this time. In reality, and despite claims to the contrary, it takes a battery this comprehensive to provide adequate measurement of the key dimensions that accurately predict successful call center job performance.

3) Make sure the selection system measures the right dimensions. Related to comprehensiveness is the question of what to measure, which is determined only through systematic research. We found the following attributes crucial to telemarketer effectiveness:

• Visual processing ability,
• Tolerance,
• Innovative thinking ability,
• Sociability and agreeability,
• Dominance,
• Perseverance,
• Orderliness,
• A strong sense for planning,
• A sense of well-being.

In our parallel research with telephone-service representatives, we found the following dimensions critical in the prediction of successful work performance:

• Verbal skills,
• Tolerance and trust,
• Visual processing ability,
• Ethical standards and a social conscience,
• Creative thinking ability and resourcefulness,
• Non-extreme, conventional values,
• Self-control,
• An absence of cynicism,
• Conscientiousness,
• A sense of well-being.

All of the above-listed key dimensions can be measured in a well-designed test battery. As seen from the above profiles, some attributes are important in both of the major components of call center work.

The Organizational Benefits Of A Scientifically Based Selection System
In our research on call center employee selection, we calculated the benefits of this type of selection system. Benefits have been indexed in two ways. Let's consider each of these in turn.

• Increase the percentage of high-performing employees.

We compared the improvement in call center job performance that results from different levels and kinds of personnel-selection procedures used in hiring. Improvement was defined as the increase in the percentage of employees who would subsequently be rated as good or excellent performers. The results, which appear in Figure 1, rest on two assumptions: (a) that the current percentage (base rate) of call center employees (not selected by a systematic process) demonstrating good or excellent performance is 50 percent, and (b) that one employee is selected from four assessed (or 10 from 40, etc.).

The values in Figure 1 represent the increase - over the 50 percent base rate - in the percentage of a call center's employees subsequently rated by their managers as good or excellent. With a scientifically based selection system in place, for example, 72 percent of call center employees would, in time, be so rated, in contrast to the 50 percent that accompanies no systematic selection.

• Yield a substantial monetary gain for the organization.

We examined the dollar yield to organizations employing systematic selection of telemarketing employees. Telemarketers were chosen because, unlike employees in many other job categories, the results of their work are easily and naturally expressed in a dollar metric. Our findings, which appear in Figure 2, again rest on two assumptions: (a) that one employee is selected from four, and (b) that the average value of a single sale is $50.

The values in Figure 2 represent the annual dollar increase per employee selected, above the normal revenues generated by telemarketing employees hired without a systematic, validated process. With a scientifically based system in place, for example, selected telemarketers can be expected to generate $31,600 per year more than what can be expected from employees selected without such a rigorous process (under the assumptions noted), and this increased annual payoff continues for every year of the employee's tenure with the organization. In a sales center that hires 50 new employees a year, this translates into a $1,580,000 increase in revenues per year from scientifically based selection.

Conclusions
By using a scientifically based selection system, a call center can expect substantial returns. With the cost of this assessment, which ranges from perhaps $150 to $350 per employee hired, representing less than 1 percent of the additional revenues that the employee selected this way will generate in one year, such a system should be seen as a first-rate investment.

Adding a structured interview to such a comprehensive test-based selection system increases the tangible benefits slightly over those depicted in Figures 1 and 2. In addition, the balance provided by an interview improves both corporate public relations and confidence among the hiring staff. For all of these reasons, we recommend that any overall hiring program include an interview to augment the information gained from a test-based selection system like the one described here.

Scientifically based selection ultimately comes down to obtaining and using all of the reliable and relevant information available to you - in a systematic and scientifically established way - to make the right selection decisions. When an organization makes this its standard practice over an extended period of time, it is laying the foundation for inevitable workforce improvement.

Ralph Hakstian is a principal of HR Decisions Ltd. and a professor of industrial/organizational psychology at the University of British Columbia. He has more than 20 years' experience designing personnel-selection systems for a variety of companies and has published nearly 100 articles in the scientific literature. Linda Scratchley, a senior psychologist with HR Decisions and a lecturer at the University of British Columbia, specializes in personnel selection, performance management techniques and creative management.

Founded in 1978 and based in Vancouver, British Columbia, Canada, HR Decisions Ltd. designs and installs focused personnel-assessment systems for selection, promotion, employee development and performance review at all organizational levels, and has particular research-based experience with assessment systems in the telecommunications industry.