With ever-faster computers and increasingly larger storage capabilities, customer interaction centers are collecting more data in more ways than ever before. CICs can collect metrics ranging from on-hold time and the number of calls in queue to average sales per call and the number of help-desk cases opened and closed. However, data collection is relatively meaningless unless the information is properly analyzed.
In many organizations, such analysis of the various touchpoints falls under the responsibility of those managing the customer interaction center, because it is often the company's focal point of customer contact. However, analysis alone isn't enough. Each company should look at its statistics in terms of its own business process, says Stephen Beckett, chief technical officer for CallCenter Technology, Atlanta.Wrap-up time
However, the relationship to the business process can't be examined using traditional call center metrics, which tend to be focused on volume measurements like calls per hour and calls in queue. "Companies have been spending too much time with measurements that are focused on volume," explains Bill Bleuel, professor of quantitative methods at Pepperdine University, Malibu, Calif. But a shift is underway. The increased emphasis on the customer has necessitated a move to quality measurements, according to Kathleen Klasnic, consultant with Datamonitor, New York.
While traditional quantitative measurements shouldn't be ignored, they need to be balanced with those measurements that reflect a company's business direction, Beckett says. So while a company needs to examine the efficiency and effectiveness of its CIC, it must pay attention to more complex issues like customer profitability.
The key efficiency metrics, according to GartnerGroup, stamford, Conn., are average monthly calls handled per agent and average monthly transactions handled per agent. The effectiveness of the call center can be evaluated by personnel costs and head count distribution among personnel categories, as well as first contact completion percentage. A poor first-contact percentage could reduce customer satisfaction and increase call center costs.
However, these aren't always the most important measures of effectiveness. For example, an insurance firm sales desk agent might need to make several calls to a prospective client before closing a sale, and each call could last several minutes. This is acceptable from a business perspective because a single customer can represent a large payoff in terms of profitability and products sold. "It's necessary for an insurance company to make people feel comfortable about doing business with them," Beckett says. "They might spend as much as 15 minutes on a single call."
For a company selling low-cost, low-margin items, however, it is necessary to make more calls in a shorter amount of time. Therefore, it's imperative that company management communicates its strategy to those in charge of the customer interaction center, say Beckett and Bleuel.
Additionally, not all measurements are necessary for every company or even for different functions within the same company. For example, a help desk rep might not perform sales tasks at many companies, while other companies might use universal (service and sales) agents. A center that handles strictly outbound sales typically won't have calls in queue. Although there has been plenty written about universal agents, most still fall into one category or the other, Beckett says. For example, Nortel Networks has three different departments within its call center, each with its own unique responsibilities (see "Making Quality Measurements a Continuous Process," this article).
Movement to Quality
While metrics have focused on hard or quantifiable skills in the past, the successful CIC will focus more on soft skills, says Oscar Alban, product consultant at Witness Systems, Atlanta.
These soft skills include items like being friendly or matching a customer's communication style, whether that's concise or conversational. These elements are impossible to judge using volume and time measurements. Alban recommends recording conversations and having staff and supervisors review them, much like a quarterback and coach studying game films.
"To get the best and most comprehensive customer data, companies are trying to use information from all customer touchpoints, including call centers, the Web and mail, so they can better tailor their products and services to meet customer demands."
According to Beckett, while both quantitative and qualitative measurements can help companies, in most cases the information isn't analyzed in a timely manner. Like Alban, he likens CIC metrics to football. Analyzing the data at the end of a week, month or quarter is like the quarterback analyzing film after a game. While that can help somewhat, it's analogous to going into the following week's game with blinders on-with everyone reacting to what happened before, not to what is happening at the time.
"You need both the tactical and the strategic analysis," Beckett says. While some in-depth analysis needs to be done periodically, automated systems also enable customer interaction center managers to see changes in real time, enabling them to react more quickly than if they wait until after the fact. He recommends that centers use automation wherever possible so that the necessary analysis can be done as quickly as possible. "Otherwise," Beckett explains, "managers fall into the trap of collecting data all week, entering it into an Excel spreadsheet, then on Monday, starting it all over again." By capturing the data automatically in a form that can be immediately analyzed, managers can quickly determine how to react to any changes.
For example, an insurance company that finds a strong increase in the number of requests for competitive quotes could find that a new competitor has entered the market. In order to profit most from that knowledge, the company needs the analysis as soon as possible, not six months after the competitor entered the market.
Projected sales of call monitoring software for the next four years show that small CICs are expected to make up a much greater share of sales as systems become cheaper and easier to integrate. (Courtesy of Datamonitor)
"The squeaky wheel tends to get the grease," says Jay Pepper, senior vice president of sales for Digital Dialogue, Southfield, Mich., a company that provides Internet and CIC support from its facilities on a contract basis.
The customer who calls a help desk 10 times a year, particularly in a low-margin business like credit cards, is very expensive to the company but generally easy to retain. The customer who has never demonstrated any relationship or loyalty is much more likely to leave if he gets a better offer. Therefore, Pepper recommends analyzing the number of calls versus customer profitability to make sure that the most profitable customers are contacted, particularly if they haven't phoned the interaction center themselves.
Witness Systems' Alban recommends that centers go one step further in monitoring the number of calls versus profitability information and actively seek the frequent customer who has stopped doing business for any significant period of time. That's the time to call the customer to make sure his needs are being met.
Interaction centers also need to look at customer profitability to offer differentiated levels of service, says Alban. The airlines have long recognized the traveler who flies 100,000 miles a year on a single airline over the flier who makes one trip a year. The airlines offer these frequent travelers preferential treatment through VIP lounges and other amenities. CICs need to do the same for their best customers. Complete, up-to-date data should immediately identify VIP callers so that they receive the best service. Several companies provide their VIP customers with a separate toll-free phone number to the CIC. Calls that can't be answered immediately go into a much shorter queue than calls to the common toll-free number.
By taking advantage of customer profitability and other metrics, companies can change the call center from a cost center to a profit center. For example, Alban previously managed three call centers of a telecommunications company. The company used metrics to segment its customers based on their profitability to the company. The company's top agents contacted VIP customers and offered them the best calling plans. Other reps spent less time and effort contacting less profitable customers, and the discounts offered were smaller than those given to the more profitable customers.
Developing the above plan based on metrics analysis helped Alban increase the company's market share for small business customers by two percentage points, while its major competitor's market share dropped by two percentage points. Each percentage point of market share was worth $600,000 in revenue.
Consider the Web
Companies are increasingly including Internet support under the CIC umbrella, according to Bob Dutile, partner in the CRM practice for Grant Thornton, Chicago. The World Wide Web creates a new set of possible measurements for CICs that have implemented the Internet.
Many of these metrics and analyses are still under development, according to Pam Berryhill, Grant Thornton senior consultant. Still, there are some early leaders, including e-mail response time and measurements of how a customer negotiates his way through a site.
Because of the nature of the medium, response times are critical. While many CICs that support Internet sites attempt to respond to e-mail within 24 hours, that might not be quick enough for Internet customers.
By examining how a customer negotiates a Web site, a company can learn what type of links or flow-through can lead to sales. If a large percentage of viewers leave the site immediately after reaching a particular spot, it could be a sign that something-a link, perhaps-needs to be changed, Berryhill says. By monitoring Web traffic and correlating it to CIC calls, a company can judge if its self-help options are decreasing the number of calls or increasing them because of site confusion, Berryhill adds. Further analysis can show how well the site leads to orders placed through the call center.
Customer Satisfaction is Top Priority
While the above measurements are important in today's business climate, the most important metric is customer satisfaction. It is generally acknowledged that it is more expensive to find new customers than to retain them. Customer satisfaction "is the most important measurement, as it can affect customer retention, loyalty and ultimately, an organization's bottom line," according to GartnerGroup.
GartnerGroup recommends that a company hoping to be more customer-focused should survey its customers about the effectiveness of its CIC. A growing number of companies have recognized the importance of customer-centric metrics and have performed customer satisfaction surveys for all their CICs, says Grant Thornton's Dutile. Among the questions to ask are:Would you transact business with the organization again?
Bleuel points out that many of the newer measurements dealing with quality are more customer-focused than the previous quantitative measurements, which were company-focused.
"Incorporating customer satisfaction into the service center measurements does not replace the need to measure the efficiency of the center," says Bruce Guptill, a GartnerGroup analyst. "Traditional measurements often provide for a reality check of the feasibility of customer service requests. In setting service levels and expectations, it is important to include both customer requests and numeric metrics-the two aren't mutually exclusive."
Of equal importance are dissatisfiers-those items that will drive the customer away, Bleuel says. Dissatisfiers are a little tricky because there are no immediate measurable benefits for removing them. Beckett explains that the customers who aren't satisfied or who get better value elsewhere will usually leave quietly. That's why outbound check-up calls are so important-they can help prevent customers from going to a competitor.
An interactive voice response unit with complex, multi-layered menus is a common dissatisfier. A customer can spend several minutes going through a menu, getting placed on hold and being transferred, yet still wind up with an agent who doesn't fill his needs. "By that time," Bleuel says, "the customer is up the wall, over the wall and ready to strangle you. Dissatisfiers are killers." He adds that the rapid growth of the customer interaction center industry has forced companies to use a bandage approach to measurement, rather than strategically analyzing the data. So while a technology may appear to reduce the number of calls that go to a live agent, the metrics need to be studied further to see if customer needs are being met.
For example, if Company A handles 15 calls in the same time period that Company B takes 20 calls, it would be easy to conclude that Company B is doing a better-at least more efficient-job. However, if all 15 of Company A's callers are satisfied but 5 out of 20 of Company B's callers were repeat callers because they were given incorrect information the first time, that assumption doesn't hold. Therefore, a superficial examination of a metric is too simplistic. Companies need to study the relationship between various items being measured. As in other technological advances within this industry, the telecom and banking markets are in the forefront of using monitoring software for their metrics analysis.
"Use qualitative and quantitative methods to determine incentive bonuses for agents and supervisors."
Using Metrics to Reward Performance
Qualitative and quantitative methods can also be used to determine incentive bonuses for agents and supervisors, Datamonitor's Klasnic says. Performance measurements, including sales, trouble resolution and quality-of-service scores can help determine bonus compensation, which can be a significant part of an agent's or supervisor's total compensation.
For example, Crestar Bank in Richmond, Va., which has a 200-seat CIC, uses metrics to monitor and compensate agent performance, says performance supervisor Ana Whitt. In a program borrowed from their sales agent counterparts who have used it for three years, service agents are expected to meet specific performance levels to score 100 percent. They get bonuses for exceeding this standard. If an agent scores 150 percent, for example, he receives his normal base pay plus a 50 percent incentive bonus. The incentive bonuses are paid out once a month, and top agents can double their salaries. An entry-level agent's base salary is $20,000, tier-two agents make $22,000 to $23,000, and tier-three agents can earn $24,000 to $26,000. Whitt explains that although Crestar's base pay is lower than what's offered by other companies in the geographic market, the incentives give agents the potential to be among the highest paid. In fact, since instituting the plan, agents have the potential to out earn some managers.
Since the new program was instituted, the turnover rate among service agents has dropped from 100 percent to 70 percent. While reducing turnover is a good objective, too drastic a change should be a cause for concern. For example, after the sales agents had been using the program for two years, there had been zero turnover. However, Whitt found that the problem was that standards were too low.
"You want to have a turnover rate of at least 10 percent," says Whitt. "The agents who are weakest are the ones who tend to leave." Logic dictates that the same would hold true for service agents-the weakest will leave out of frustration.
Not Too Late to start Reaping Benefits
It appears that a growing number of companies plan to increase data collection and analysis in an effort to retain agents and customers and recruit new ones. A study by META Group, stamford, Conn., forecasts growth in the business intelligence population-suppliers, business partners and consumers-from 400,000 in 1987 to more than 10 million by 2010. According to AMR Research, total aggregate revenue of the customer relationship market, of which data warehousing is a part, will grow from $1.2 billion in 1997 to $11.5 billion in 2002.
The Datamonitor report titled U.S. Call Center Software Markets pegs the call monitoring software market as "one of the most attractive sectors of the call center software industry. Worth $120 million in 1999, the value of the market will increase more than seven-fold by 2003 and contribute $850 million to the global call center software industry. Although, at present, market share is tightly consolidated among the sector's top three vendors, Datamonitor believes vendors best able to concretely improve upon existing technologies and effectively market innovation will certainly be able to wrest market share from the leader." Such a prediction can only bode well for both vendors and their potential clients as the collection and analyses of new metrics become increasingly vital to one-upping the competition.
Since the quality of customer service is a vital way to differentiate companies from each other, implementing a solid program of analytical metrics-and acting on the results-can be the factor that helps separate the ho-hum from the gung-ho.