Tiny Tweaks Improve Testing Results

Customizing prehire assessment tests to achieve specific business needs can be a long, complex process, but even small adjustments can improve results.

February 22, 2010

If you’re wearing sexy sunglasses, they were probably made by Luxottica Group, the world’s largest eyeglass company. The Milan-based firm manufactures and retails 12 house brands, including Ray-Ban, Vogue and Oakley, and 22 licensed brands, including Dolce & Gabbana, Versace, Prada and Chanel.

The worldwide workforce of 61,000 includes 35,000 employees based in North America, where Luxottica hires 12,000 to 15,000 new employees a year. Ninety percent of the new hires are retail associates, and all of them take a 20-minute prehire assessment test at a cost of $9 to $10 per test.

Carol Spicer, vice president for human resources administration at Luxottica Retail headquarters in Mason, Ohio, believes that the investment in testing is fully returned to the company within an employee’s first hours on the job. The company recently completed an assessment test return-on-investment study for its 470 stores in the Pearle brand. The study found that new hires who scored in the top quartile on the assessment test sell an average of $14 per hour more than those who score in the bottom quartile.

On an annual basis, those results translate into an additional $18,000 in sales for each employee scoring in the top quartile. If the company could fill all Pearle sales associate positions with employees scoring in the top quartile, it could boost annual sales by $12 million, according to Spicer.

Luxottica has achieved this recruiting intelligence after more than a decade of systematically administering, evaluating and tweaking its prehire assessment tests to achieve specific workforce objectives. The company has pushed the customization process to a highly advanced level, but even minimal customization can produce solid results for many employers.

Test, evaluate, modify

Assessment is one step in Luxottica’s 10-step recruiting process, which also includes competency-based interviews. The company’s legal team reviews hiring every quarter to ensure equal opportunity. The assessment test cutoff score for hiring varies based on the legal reviews. “If you make the funnel too small, you can’t hire enough people,” Spicer says.

Luxottica started with an off-the-shelf standard test from PreVisor, a testing firm based in Atlanta, more than a decade ago. “Luxottica provides a perfect example of the optimization process for fine-tuning tests to meet business needs,” says Ken Lahti, PreVisor’s vice president of strategy and content.

“The point is to determine how well the test is predicting employee performance,” Lahti says. “Then we can make improvements by adding or removing content or changing the scoring to make it even more predictive in the next round of testing.”

To customize a test, PreVisor uses job analysis to determine what should be measured. “We devise the customized test and the employer administers it and hires on the basis of the scores,” Lahti says. “New hires represent a range of scores because no employer can access only top-scoring candidates. Then we can track the employees over time for performance, sales and shrinkage.”

In addition to documenting higher sales at stores with larger numbers of higher-scoring employees, Luxottica found that employees receive better performance reviews and shrinkage declines.

“Shrinkage covers all forms of inventory loss, including customer and employee theft and losses resulting from employee carelessness,” Lahti notes. “Employees have some control over all aspects of shrinkage. Shrinkage is a good metric, but it can only be measured on a storewide basis. An employer can compare the overall test scores for employees at a store with the store’s shrinkage to determine if there is a correlation. Luxottica found that there is one.”

The 2009 Pearle brand return-on-investment study found that shrinkage is lower at stores with larger numbers of employees who scored in the top quartile on the prehire assessment test. Shrinkage averaged $19,000 per year at Pearle stores with larger numbers of higher-scoring associates, compared with $22,000 per year at stores with lower-scoring associates.

Luxottica also uses a customized assessment test for managers at its 850 LensCrafter stores. At the end of the five-quarter tracking period, a study showed that managers who scored within the recommended passing zone moved 7 percent more products and services than those who did not reach the recommended score.

Ongoing tweaks

Because Luxottica is a large company with a long history of testing, the company has a rich database and PreVisor can modify its tests based on company-specific data.

“We can tailor the metrics for Luxottica, make any modifications necessary to accommodate changes in jobs or the business model, and provide recommendations for further customization,” Lahti says. “For example, Luxottica changed to a more consultative sales model and this had implications for the assessment test.”

“You never have a perfect assessment test because businesses change,” Spicer says. “All of our brands continue to change, so we need to tweak the tests. At LensCrafters, for example, we needed to add a diversity component to determine if applicants understand the diversity of the customer base. We reassess every brand every two to three years to see if we need to tweak the test.”

Luxottica also tweaked the test for its Sunglass Hut brand, a particularly tough brand for recruiting and retention. With 8,700 employees, Sunglass Hut accounts for 25 percent of the total Luxottica North American workforce but 45 percent of all new hires. Turnover at Sunglass Hut is well above the 35 to 40 percent turnover at other brands and double the 37 percent companywide turnover rate for North America. At the end of 2009, Sunglass Hut turnover stood at 73 percent, down from 110 percent in 2002. Spicer believes the decline is largely the result of additional customization in the prehire assessment test.

One modification in the Sunglass Hut test stemmed from the brand’s use of kiosk stores, where an associate often works alone. “We added a ‘culture fit’ component to the assessment, with questions about whether candidates like to work on their own, which means they might be a fit for Sunglass Hut and we can steer them in that direction,” Spicer says. “We found that adding the culture fit questions improved retention. People stay on the job twice as long.”

The amount of customization required to create optimal assessment tests varies by position.

“The axiom in assessment is that more is better,” Lahti says. “From a scientific perspective, we would like to conduct three-day assessments that measure everything, but there is a tradeoff with the client’s needs and the applicants’ time. Luxottica has been able to fine-tune through field experience, so a quick test can produce good predictions.”

“We want to go out the gate with assessments that work,” Lahti says. “The tests are already customized at the competency level. Even one round of testing, evaluating and modifying will be very powerful.”

Workforce Management Online, February 2010 -- Register Now!