HR practitioners often struggle to find methods to prove that their programs are effective and thus worthy of budget dollars. The author of this article recommends using “split samples” to test the effectiveness of a program. A new program is implemented with half of a certain employee population and withheld from the other half, and then performance is measured after a period of time has passed. The theory is that after exposure to the HR program, performance should improve within the control group.

For example, if a company has 30 sales people in a region, 15 are selected at random and put through a training program. The average sales of each salesperson are documented prior to the training and then again six months after the training, to see if there has been improvement.

Other areas can benefit from this type of split-sample measurement. In recruiting, the employee referral bonus program can be reenergized in one unit and not in another to see if the program results in better quality hires. In executive coaching, half of the executives can be coached and then 360-degree assessments and business results can be analyzed to see if this group achieves better results than the uncoached group. In compensation, team incentives can be added to the compensation mix in one facility or region to see if business results improve compared to other locations that do not receive incentives.

Most senior managers will demand hard data to support the premise that a program works, no matter what the discipline. It can help to involve the CFO’s office, representatives from the Six Sigma team, and market research and departmental managers early in the split-sample selection process. This can help to ensure that any complaints or concerns about how measurements are to be attained can be handled from the outset.