How do you gauge success in HR service delivery? Operational metrics, such as response time and speed to answer, are not enough, according to Christi Rager Wise, the Global Service Leader for Customer Experience at Hewitt Associates. "We can measure how well we've adhered to operational standards, but the ultimate litmus test is how customers feel about their experience," she says. "If HR service delivery isn't meeting customers' needs, you won't be able to influence them to adopt behaviors that achieve business goals."

Hewitt has long been committed to ongoing research to ensure the optimal customer experience, one that anticipates customers' needs and helps them make the best use of their HR and benefits resources. "We can't impose an experience on the customer," says Wise. "It's the customer who owns the experience. But we can gain insight into how to optimize it. Our commitment to research is an outgrowth of our commitment to the customer."

What Drives Satisfaction?
Hewitt has conducted extensive research to identify the drivers of customer satisfaction with the Web, service center, and print communications. "Overall satisfaction scores don't give you the insight you need to take corrective actions," says Wise. "It's not enough to know your score; you need to know why you got that score. And that requires knowing the satisfaction drivers for all service delivery channels. They're the most important lever we have to impact customer satisfaction and influence behavior."

Through focus groups, surveys, pilots, and statistical modeling, Hewitt has identified key satisfaction drivers for each channel. Some drivers, such as "ease of navigation" (Web) and "arrives at the right time" (print), are channel-specific. Others, such as "first-call resolution" (service center) and "answers all questions" (Web), are conceptually similar.

The research also revealed that not all drivers are equal. "Some drivers, such as 'ease of access to the Web,' are 'must-have' factors," says Lisa Miller, Hewitt's Customer Experience Measurement Strategist, who designed the research. "If employees feel they have great access, that won't boost satisfaction scores, but if they perceive that access is poor—for example, if they have a poor connection speed—scores could decline. It's not a linear relationship."

In addition to the must-haves, researchers identified other aspects of service delivery that exceed customers' expectations and can take satisfaction to a higher level, such as helping them "make the best decision possible." "These variables have a nonlinear relationship in a positive sense," says Miller. "Their absence may not lower satisfaction scores, but their presence can significantly boost satisfaction."

For Hewitt, total satisfaction is the benchmark. "We think our customers' expectations are entirely reasonable, and we're committed to making customers completely satisfied," says Wise. With that in mind, Hewitt queries any customers who rate their Web experience as less than totally satisfying, to learn why—and how Hewitt can earn the highest score.

What it takes to earn the highest score changes over time. "Customers keep raising the bar," says Wise. "In the past, we could get high satisfaction scores by making it easy to complete a Web transaction. Now, to be completely satisfied, customers must feel that their experience helped them to make the best possible decision, and one day that could be a basic requirement."

To keep pace with customers' rising expectations, Hewitt measures satisfaction on a regular basis. "We always try to stay a step ahead, to anticipate what it will take to completely satisfy customers tomorrow, so we're prepared to deliver on those requirements today," says Miller.

Applying Insight
Hewitt uses the insight gained from satisfaction research to identify and prioritize enhancements. Satisfaction drivers also provide a framework for problem resolution. "We know exactly where to look if our satisfaction scores slip," says Wise. "Our research shows what's most important to customers, and that's where we focus our attention."

Satisfaction research also provides insight into how channels should be deployed to enable the best customer experience overall. For instance, 75% of defined contribution interactions involve balance checking, and for this activity, the Web is the overwhelming channel of choice. But the choice changes, according to Miller, depending on the customer segment and task being performed.

For example, in 2005 Hewitt interviewed 350 recent retirees about their experience with the retirement process. "Three-quarters of these customers used the Web during the retirement process," says Miller. "They especially liked the modeling capabilities and the fact that they could access the Web 24/7—running pension estimates in their pajamas if they wanted to. But they also made it absolutely clear that they wanted access to a person who specializes in retirement. These experts play a unique role in the process by providing guidance and building confidence. Through this research, we learned that this combination of channels is necessary to give our customers the best possible retirement experience. It's a good example of how you can't impose an experience on the customer. You have to follow their lead."

Measuring Customer Behavior
In addition to assessing satisfaction, Hewitt measures customer behavior to glean insights that help guide Hewitt's solution development strategies. "We conduct two types of research to learn more about customer behavior," says Wise. "Sometimes we mine Hewitt's database of nearly 20 million people. For other experiments, we start with a hypothesis and test it on small groups of customers, using statistical sampling methodology." These frequent, quick experiments elicit statistically valid results that allow Hewitt to constantly monitor and react to what they're learning.

Hewitt is currently testing the hypothesis that employees given the option of paying for health care premiums via automatic credit card transactions would be more likely to stay in the plan than those who write monthly checks. "We want to know not just if credit cards make a difference, but if the difference is big enough to justify the cost and effort to develop the capability to handle those payments," says Wise. The results will help shape Hewitt's strategy in this area.

Another research effort, conducted by the Industrial Engineering Group at Hewitt, addressed the optimal length of the annual enrollment period for health benefits. "We predicted that more people would enroll close to the deadline than near the beginning of enrollment, since it's human nature to procrastinate," says Wise. Testing confirmed that transaction volume peaks in the last week, regardless of the length of the enrollment period.

"It's impractical to vary staff levels each week, so customer service teams must be staffed to handle peak volume during the entire enrollment period in order to maintain the required service level," says Miller. The longer the enrollment period beyond a two-week baseline—or three weeks for high-volume clients—the more it costs to handle the same number of calls. The cost implications are significant, as the chart below illustrates. Validating this hypothesis enabled Hewitt to help clients optimize enrollment periods while controlling costs. Hewitt also uses that information to stagger enrollment periods, so that one client's valley is another one's peak. "This enables us to meet our clients' expectations while operating with greater efficiency," says Miller.

Informing Decisions
Measuring customer satisfaction and behavior provides directional input. But to help employees make the best possible decisions during benefits enrollment requires digging deeper. "Knowing that we're evaluated on helping employees to make good decisions doesn't tell us how to help," says Wise.

To find out, Miller surveyed hundreds of employees who had enrolled for their medical benefits and found that five "C's" drive enrollment decisions: cost, coverage, changes (since the last enrollment period), comparison (of plans), and confirmation that the employee's doctor is in the plan. "Now, whenever we communicate with customers about benefits enrollment, we make sure to prioritize these factors, and de-emphasize anything that's not related to them," says Wise. For example, her team integrated the factors into the most recent version of Your Benefits Resources^TM. They redesigned the home page, giving visual prominence to cost, coverage, and comparison; expanded the physician search capability; and reduced the number of links unrelated to the five C's.

The team is also helping customers make good decisions in other areas of employee and manager self-service. "When managers initiate a work-related event for one of their employees, MyHR® [Hewitt's desktop application for workforce administration] presents a list of all the items they should consider or act on," says Wise. "Suppose a manager is transferring an employee to another location. Will there also be a change in role or in pay? Is relocation assistance offered? We provide all the information they need to consider, next to links they can click on to take action immediately or to seek more information."

From Discovery to Prediction
Helping employees make good decisions will become even more important as companies take the next step toward total benefits decision making, in which both employees and employers seek to maximize the value of their contributions across all benefit programs: health, welfare, financial security, and retirement. Empowering employees to make decisions in the context of their total benefits is a powerful concept, but it adds complexity. To help employees make complex choices, Hewitt is exploring online modeling tools that offer suggestions tailored to an individual's lifestyle and financial situation. Another potential approach: presenting employees with two or three prepackaged alternatives, tailored to their needs. "We are committed to doing all we can to make the complex simple," says Wise. "That's where we can add the most value."

Looking Ahead
With each new survey and experiment, Hewitt acquires more data to help in predicting customer behavior. "We're applying our collective knowledge and experience, based on measurement, to make more accurate predictions as we encounter new situations, services, customers, and businesses," says Wise. "We'll never stop learning from our customers, gaining insight into how we can serve them better."

 Back to Top