Re-Evaluating Performance Evaluation

When Lou Gerstner took over as CEO of IBM in 1993, he requested to see the customer satisfaction scores for the database division. The results: 87% satisfaction. That’s a decent score, until you consider that the customer attrition rate over the previous two years was about 66%. IBM was measuring the wrong things. And, we can reasonably assume that IBM was basing evaluations and employee compensation on the wrong things.

A common trap in companies is allowing silos to work from isolated metrics specific to their own function—and we’ll explore this concept in detail in Chapter 12, “Measure What Matters.” One impact is the downside of employees doing what you pay them to do, particularly if they are evaluated on their performance against inappropriate metrics. If you measure and compensate them on internal issues only, they will focus on internal issues only.

If you don’t connect employees to customer experience in determining compensation and bonuses, you’re not going to get results. Or you won’t like the results that you are getting. You will see two-thirds of your customers that are 87% satisfied looking for greater satisfaction with a competitor. That means that you must begin holding the employees accountable for the ultimate performance of the company, and not merely for some internal measure.

We were auditing an airline service operation, and chose to concentrate on analyzing the experiences of upper-tier members of the airline’s frequent-flyer program. These customers have flown at least 75,000 miles in the previous year, so they certainly represent the carrier’s most loyal customers, and presumably the most profitable customers. My intent was to find the best examples of customer service given to these special customers that we could apply back to those in the lower tiers of the program. During the audit, I noticed something very interesting. Watching the monitors that tracked incoming call traffic, I saw that the call system would cut off customers who were on hold for more than 59 minutes. You read that correctly—not 59 seconds, but 59 minutes. I was stunned to learn that the cut-off was exercised on 400 customers a week. Each week, 400 of the airline’s best customers were subjected to ridiculously long hold times, and then simply dropped from the system.

When I talked to the VP of customer care—who, by the way, began the conversation asking if I’d learned about any “quick wins” they could deploy immediately—I told him that a configuration error seems to have crept into the system, cutting off the most- profitable customers. He said, “That’s not a configuration error. That’s how I manage the system. It’s very simple. I’m being measured on average handling time, which consists of wait time and talk time. If we answer 400 people within 59 minutes or more of wait time, I’m not going to make my numbers. My kids are not going to make it to college.” He paused, then said, “By the way you caught me on a good week. Sometimes I cut them off after 25 minutes.”

“What about the CEO’s number?” I ventured. “The one in the memo from the top office that declares, The customer is absolutely number-one.”

The VP said, “I have no idea what you’re talking about. I manage a single number: average handling time.” The VP was leveraging a little-known customer service secret. Customers who drop off voluntarily—”abandoned from hold,” in the jargon—aren’t calculated into average handling time. If customers refuse to listen to the generic music and the persistent announcement that “Your time is truly valuable to us” over and over, that’s the customers’ problem. They’re not patient enough. And as my smart-aleck VP of customer care told me, “I just turned it from a reactive situation from to a proactive one. I didn’t want customers to wait so long until they finally decided to hang up themselves—I was just trying to help them out.”

If you want to start a revolution and change, re-evaluate the components of performance evaluation. And a revolution it is indeed, because productivity-based or efficiency-based evaluation and compensation conflict with a customer-centric strategy.

Install customer-centric goals related to each component of performance evaluation, including:
• Job descriptions. Do Key Performance Objectives stress functions or results? Have you formalized customer-centric objectives formalized, or have you presented them as just “good ideas”? People do what they are being paid for, not what they are told.
• Metrics. Are employees’ numeric goals function-centric or customer-centric? If you don’t hold them accountable to ultimate customer spending, they won’t care about that and won’t work to impact it.
• Compensation. Employees do what you pay them to do. You pay them to engage in fragmented activities, they’ll be happily fragmented all day long. If an insignificant portion of their pay depends on customer results, customer results will get insignificant attention.

The more-advanced companies today use performance evaluations as a coaching mechanism and not as a compensation benchmark. The coaches present the criteria that all employees should be evaluated against, and design coaching based on performance against those criteria. These companies don’t measure purely on how many servers IT fixes, but also on how many customers benefited and how the company’s bottom line benefited from fast, uninterrupted server performance.

Instead of traditional task-list evaluation criteria, install job descriptions, metrics and compensation that reflect performance against these criteria:
• Decrease in customer complaints
• Improved customer’s perception of and attitude toward the brand/organization
• Increase in customer “Thank You” letters
• Increase in customer ideas and insight
• Decreased customer attrition rate
• Increase in customer referrals
• Increase in overall customer base
• Decreased cost of new customer acquisition
• Increase in customer interactions with the organization
• Increase in customer and prospect purchase consideration
• Increase in customer acceptance
• Increase in customer upgrades
• Increase in sales of accessories
• Increase in cross-selling results
• Increase in customer overall wallet share
• Overall growth in business per customer
• Increased annual customer value
• Increased lifetime customer value

Customer-centric factors must be included in all departments in the organization. Customer experience attributes, combined with touch point analysis, will assist in implementing an organization-wide customer focused plan, as we’ll see in Chapter 12.

On top of this, look to clarifying standards for performance recognition and Moment of Truth recognition (which we’ll discuss in a moment), as well as change recognition and incentives. By “change recognition,” I refer to incentives to facilitate the change management needed as you put the customer experience program into place, to be effective over a stated period of time during the change. You want to reward those employees and departments that adopt the new program quickly and willingly to send a message to the organization as a whole. We cover this topic in greater detail in Chapter 13, “Leadership and Change Management.”

As a general rule, base 30-35% on the sorts of joint goals listed above—nothing less than that. I’ve seen plans that tie only 2% of compensation to customer satisfaction and customer performance. Who’s going to sweat for 2%? Who’s going to slow down their rush to happy hour for such a slim incentive? The remaining 65-70% is then tied to the employee’s specific function.

American Express, for example, ties 70% of call center personnel compensation to an annual customer-satisfaction survey that asks customers to rate not the department or the organization, but the specific agent the customer interacted with. Amex employs this system to manage 15,000 agents in the U.S.—1.5 million complete surveys that they can attribute directly to employees.

Scores are used as benchmarking not only for compensation, but also for coaching. Employees are informed that their supervisors are there to coach, and their supervisors are on their side. And indeed they are, because supervisor compensation is also linked to customer satisfaction, in increasing percentages at higher levels of management.

The survey is based on Net Promoter Score metrics, measuring customer willingness to recommend the agent based on the customer’s experience. The survey measures how the employee handled the call, and recognizes that the problem that motivated the call might not allow the agent to achieve first-call resolution. Therefore, the measurement focuses on call handling and not issue resolution, seeking to understand if the agent has handled the customer with such tact and service that the customer comes away with a memory of a great customer experience with that particular touch point.

In fact, one of the best tests of your readiness for customer experience lies in your employee compensation. If you pay people based on productivity, you get productivity. If you pay them based on quality, you get quality. In our Customer Experience Management Benchmark Study, just 31.3% of executives say their compensation directly reflects the company’s commitment to customer experience, meaning that in well over two-thirds of the cases, employees are ultimately getting paid to do little more than execute processes.

Just 40% of executives say that their employees have the tools and authority to solve customer problems and to do what they need to do for the customer; 38% report that employees are well-versed in how to delight customers; 31% say that current compensation reflects their company’s commitment to the customer experience. Understand what these numbers are telling us: Most employees show up to work with nothing more than a pathetic smile to serve the customer with.

And note, as well, the conflict that these numbers show. The number of executives confident that their companies deserve employee loyalty is rising, while the number who provide adequate tools to those employees is shrinking.

An excerpt from the upcoming book Customer Experience Strategy – The Complete Guide From Innovation o Execution (4i, 2010).