Ask a manager who his or her top performers are and you will usually get a quick answer. Ask for the basis of that selection, however, and you may be surprised how fuzzy or imprecise the criteria are. It is not unusual to discover that managers have no measurable, repeatable criteria for identifying this most important group of employees!

In some jobs, measurable and objective criteria are relatively easy to identify. A salesperson, for example, may be measured on sales production (units, dollars, profit, etc.). Appropriate measures for sales performance may also include calls made per unit of time, customer service measures or revenue growth. Performance in other jobs may not be as easy to measure.

• What are the metrics for a top performing social worker?
   
• How do we measure performance of a researcher in a pharmaceutical company where 20 years of research by 50 people may go into the development of a single new drug?

Often, faced with the task of identifying performance measures in a field with soft outputs or very long-term outputs, managers fall back on personal likes and dislikes, personality conflicts/lack of conflicts or other measures with little relationship to their company’s goals, profitability or long-term success.

In some settings, recognition of these challenges result in attempts to make evaluation more objective. A favorite tactic in these settings is supervisory rating scales, where supervisors rate incumbents on one or more dimensions thought critical to performance on a numerical scale that may run from three to 10 points. The outcome of such ratings may look objective, as we tend to associate decimal numbers with objectivity: “She scored a 2.7 of a possible three!” Unfortunately, analysis of range compression (where everyone scores in a one-point range of a possible three) and interrater reliability raises serious questions about the validity and utility of these procedures.

If the procedures are flawed, the outcomes will be equally flawed, with serious consequences for the business and the employees affected.

What can a manager do to avoid these pitfalls and select top performers on the basis of objective, repeatable and predictive criteria? Fortunately, since the issues in selecting top performers for job fit assessment are essentially the same as those surrounding the entire topic of performance appraisal, the literature is rich with sources offering guidance.

Interested readers might start with these titles:

• The Complete Idiot’s Guide to Performance Appraisals, Adele Margrave and Robert Gorden, 2001.
• Performance Management, Robert Bacal, 1998.
• The Performance Management Activity Pack, Terry Gillen, 2001

As the current data set demonstrates, similar savings and improved performance can apply to credit unions as well. Volume two, Issue two, featured a mid-sized credit union whose preliminary data illustrated that their strategic assessment program was achieving desired results. After a full year, the early indications are confirmed.

The credit union adopted a funnel model of selection. At the wide end of the funnel, applicants are screened for suitability on the basis of their application documents. Those chosen to enter the interview process first complete an honesty-integrity assessment, the Step One Survey II™ (SOS2). With a strong applicant pool, the credit union applies a relatively high criterion to the scores on that instrument. The criterion, combined with an initial interview (using the assessment’s interview guide) selects approximately 40 percent of the pool to continue the process. At this point, candidates remaining in the pool complete a job match assessment specific to customer service jobs, the Customer Service Perspective™ (CSP). If their match to the success pattern for the job under consideration is favorable, they also complete a job match assessment specific to sales, the Profile Sales Indicator™ (PSI). A final interview is conducted considering the complete file of information available (assessments, employment history, reference checks and interview results) and a job offer decision is reached.

After a full year of this effort, the evidence is clear: the program works as designed. Results of the turnover reduction are illustrated in the graph below. More importantly to the credit union and its owner-members, return on their investment is exceptional, and will allow them to continue pursuit of improved customer service.

The baseline group for this case study consisted of the last 100 employees hired prior to the beginning of the assessment program. In this group, failure rates were compiled for 30, 60, and 90-day cumulative failure rates, as well as total failures over the study period. Failure was defined as leaving employment with the operation for any reason.

Beginning with the study period, every applicant selected as eligible for an interview completed the Step One Survey II™, an honesty-integrity assessment which measures attitudes on four scales:

Integrity, Substance Abuse, Reliability, and Work Ethic. The SOS2 also includes a measure of distortion—exaggeration in the positive direction.

Hiring procedures were modified and a criterion level was adopted, based on research using the SOS2 assessment in similar employment settings. A distortion score of one or two, or any two scale scores of three or less, were considered a negative factor in the total employment decision, much as a negative reference or unexplained gap in work history would be. Managers who chose to consider a candidate who scored below criterion level could do so and could hire, but were required to provide reasons for their decision. The interview guide produced by the assessment was used in the interview to investigate any negative indications in the results.

Over the six-month study period, 302 assessments were administered and 155 hires were completed. The relatively high 50 percent ratio reflects the shallow nature of the applicant pool in this setting. Presumably, managers were forced to balance the desirability of individual candidates with the necessities of the operation and the reality of the pool.

Further indication of the dilemma managers faced is provided by the nature of the failures—less than 10 percent of the people who left were under involuntary terms.

Results of the study are summarized in the table below. Surprisingly, 30-day failure rates nearly doubled with use of the assessments, but overall hire failures dropped in every other category and were reduced by 33 percent overall.

Based on a cost-per-hire of $3,000 (frequently used in the hospitality industry), this represents an 855 percent return on investment.