Discovery Surveys, Inc.
Specializing in Employee Opinion and Customer Satisfaction Surveys
Improving the Workplace

By Bruce L. Katcher, Ph.D. President, Discovery Surveys, Inc.

One third of employees intend to leave their jobs.

Two companies approached me within the last few months with similar problems.

A beverage company was perplexed that it has been losing its sales reps in large numbers. The company has a great product, pays their sales people handsomely, and treats them well. It recruits talented business majors straight out of some of the best liberal arts colleges in the country. It invests a great deal of time and money orienting new talent to the firm and training them. But more than half of new employees leave within two years. The company is pouring money down the drain and sales have been steadily declining.

A small boutique consulting firm is concerned that the new younger consultants it has been hiring do not seem to be highly committed to the firm. It hires MBAs from the best Universities and are able to pretty much guarantee them a 6-figure income within the first year. The firm bends over backwards to make these young, talented employees feel comfortable in the culture. But several have recently left after only a few months on the job and management fears many more are considering leaving. They wonder whether the attitudes and values of the new generation Y employees are clashing with those of the established boomers who are the senior consultants and managers in the firm.

Representatives from both of these companies told me that they have got to stop the hemorrhaging. Losing these employees is costing them time, money, and the precious goodwill of their customers. They asked, “Is there any way we can reliably predict before we hire them who will stay with us and who will not?”


Turnover, especially of younger employees, has become a major problem for organizations. There are, of course, many reasons why young employees leave. In this newsletter, I will focus on making certain you hire the right people to begin with.

When hiring employees companies need to be able to reliably predict:

  1. Do they have the skills and abilities to succeed?

  2. Will they fit in to our culture?

  3. Will a job in our company be consistent with their personal goals (e.g., money, advancement, prestige, lifestyle, etc.)?

  4. Will they stay?

These are very difficult questions to answer precisely, even for the most experienced interviewers.

Most organizations have been recruiting at the same locations, and using the same job application, interview questions, and hiring procedures for many years with mixed results. There is no reason to expect that things will get any better without changing any of these practices.



There is an objective, scientific, and relatively accurate approach which research has shown is generally more precise in predicting job success than testing, interviewing, or any other single approach to hiring. It is called "the biodata approach."

What I am about to describe is an approach that can be effective in helping organizations improve their ability to hire people who will both succeed and stay. Wouldn't it be nice to become aware that there is information you can gather during a job interview that can reliably predict who will leave in the first year and who will stay much longer?

A few cautions: 1) Statistically, this approach requires a large sample size. 2) This approach is best used in combination with other hiring methods (i.e., tests, job interview, and references). 3) Although this approach is typically much better than most other methods, predicting job success and retention is not a perfect science. 4) It is an important imperative that you not use this approach to discriminate against prospective employees due to their race, religion, ethnic background, gender, national origin, sexual orientation, or disability.

  1. Create a Database of Potential Predictors

    Examine your personnel files and enter in to this database every conceivable piece of information about each current and former employee you hired in the past several years. This information should include every detail from the job application, notes from job interviews, information on their resumes, scores on any tests you may have used, etc. Also include anything else that you now know about these current or former employees (e.g., their life experiences, work experiences, interests, values, beliefs, attitudes, personality, personal

  2. Include Measures of Success in the Database

    Add to the database all information about their job performance that you would like to accurately predict such as performance ratings, sales, productivity, attendance, measures of work quality, and teamwork. Then add information about whether they are still with the organization or how long they stayed. These are the criteria that you want to predict.

  3. Correlate the Predictors and the Success Factors

    Conduct statistical analyses (i.e., correlation and regression analyses) to assess the relationships between the predictors and the criteria. (You may need a consultant to help you with this step in the process.) You are seeking to find predictors that are highly correlated with the criteria. If the variables are highly correlated, one can be used to predict the other. Regression analyses can determine the combination of variables that can best predict the criteria.

  4. Use the Results to Revamp Your Recruiting and Hiring Procedures

    The best predictors should now be included in the information you gather during the hiring process. Some of the predictors may be questions that you will now have your interviewers ask job applicants. Other predictors might take the form of items you would include in the application form or in pre-hire tests. More commonly, a biodata questionnaire is constructed. (Here you may need the help of a consultant as well to help you develop and validate the biodata questionnaire.)

The results of the study may challenge your long-held assumptions about the best way to recruit and hire new employees. For example, let's say you find that those you hired from Ivy League colleges don't last long. You might want to skip the Ivies and start recruiting from different schools.

The results of the study may also change how you weight different information about job applicants. Let's say that your study reveals that those with the highest college grade point average don't last long. In that case you might want to de-emphasize college GPA when reviewing job applications. Or, you may find that those with no previous experience in your field stay longer. You might then limit the recruitment of experienced workers from other companies.


The biodata approach is based on the principle, "the past is the best predictor of the future." It can help you abandon selection practices that are only moderately successful by forcing you to focus on information about job applicants that is truly predictive of how well they will do on the job.

Numbers don't normally lie. Use this proven, systematic, empirical approach to help you hire employees who will both succeed and stay longer with your organization.


All material is © copyright , Discovery Surveys, Inc.,