Spending HR Dollars Wisely
Given the state of the economy, HR executives are being forced to make increasingly difficult decisions regarding their current and future expenditures. And with open enrollment season right around the corner, many are wrestling with the correct mixture of benefit offerings that are right for their population. This brings us to our guest post by Carol White of Sawtooth Technologies. In the spirit of transparency, it’s important to know that the only relationship between Inflexion and Sawtooth is a mutual desire to further our industry. Please take the time to read her thoughts and - as always - share your comments and experiences. Take it away Carol!
In HR, do we think of employees as our customers? Should we? When it comes to spending our company’s money, it’s our job to maximize employee satisfaction for each dollar we use.
But, how do we know what employees want? We know about the people who complain to us, and maybe about the people who share compliments when we change a benefit, but how can we be more systematic and intentional about maximizing satisfaction per dollar?
Many companies conduct routine surveys about employee satisfaction, perhaps even having employees rate benefits they are considering offering (or reducing) in hopes of finding out what matters most. But, often, the results produce ties… people want it all. As we face decisions about what to offer, we must make trade-offs: Should we offer better health coverage or a richer 401(k)? Should our health plan include a lower co-pay or a lower out-of-pocket ceiling? If costs are the same, which would contribute more to employee satisfaction and retention?
We can do a better job at making these difficult and costly decisions by having our customers (i.e., our employees) make the trade-offs. Research techniques that are most often employed in product and service development for end-users can also be used with employees. So, instead of having to guess at the trade-offs, we can measure them.
Two techniques - conjoint analysis and maximum difference scaling (”MaxDiff”) - are used to measure such trade-offs. Although we cannot give the tutorial on these techniques here, we can provide a short introduction and offer some links.
Conjoint Analysis
Conjoint analysis is often used in service design. It asks customers (or employees) to select a favorite offering from a set. Then, it changes the elements of the offers over and over, continuing to ask for favorites. For instance, employees might need to choose between three hypothetical health plans:
- $50 co-pay for routine visits; $20,000 maximum out-of-pocket; and no extra services
- $25 co-pay for routine visits; $50,000 maximum out-of-pocket; and vision and dental
- $75 co-pay for routine visits; $10,000 maximum out-of-pocket; and dental
As you can see, the employee has to make a trade-off and cannot have the best co-pay, out-of-pocket, and extras. Through a series of similar questions that would include many other health plan features and trade-offs, we can analyze each person’s responses and quantify his/her value system - how much each level of each benefit is worth. Then, we can predict the person’s preference for any combination - any health plan configuration. And, finally, we can combine all responses from all respondents and use a predictive simulator to identify winning combinations that maximize satisfaction/preference at set cost thresholds. (Click here for more details.)
MaxDiff
In MaxDiff, the employee also makes trade-offs, but they are across individual benefits rather than the combinations of benefits seen in conjoint analysis. A MaxDiff study starts with the list of items we want to prioritize, such as potential employee benefits. From that list of items, we just show the employee a smaller subset of items at a time and have them choose the most and least important in that subset. For instance, a question might ask an employee which is the most and which is the least important aspect of their benefits from a list like this:
After asking a series of these questions with different mixes of benefits, and by analyzing patterns of response across the subsets, we can calculate a MaxDiff score for each item, representing the relative importance of each. Benefits that might make sense to take into a MaxDiff could be those which, at a high level, would help us determine which aspects of employee packages are most worthy of investment: retirement plans, health insurance, life insurance, flexible hours, work-from-home options, etc. Alternatively, they might be at a micro-level, such as which benefit “extras” are most valued: free soda, company-provided Friday lunch, monthly parties, 15 minute neck and shoulder massages at lunchtime, weekly yoga classes, etc. (Click here or here for more details.)
Leading organizations in financial services, temporary staffing, high technology and more have used these advanced research techniques successfully to model employee preferences for health insurance, retirement plans, and compensation package elements. They have used the results to focus their energies on the aspects of employee benefits that are most important to attracting and retaining talent, to allocate resources wisely, and to select benefits packages that provide the best balance between employee satisfaction and cost.
Carol White has more than 25 years of experience in conducting market research on preferences and needs. Currently, she is a Partner in the Sawtooth Technologies Consulting Group practice, using her expertise to help a wide variety of clients use market research and information to make strategic decisions and take action. You can follow Sawtooth Technologies on Twitter or contact Carol directly at cwhite@sawtooth.com.