One particular challenge of managing call centers seems not to have changed much in the last year. You're so busy running a call center that you barely have time to evaluate it. Yet when you do find the time to review your center, the data that's most easily available to you is only useful for short-term decisions.

This challenge has, to a limited extent, a technological solution. In this article, we're emphasizing analytics tools, which refer to software that provides call center managers with real-time and historical information about how efficiently agents handle calls. Such tools are useful for tracking whether your center fulfills a minimum requirement that agents be available when customers try to reach them.

We've divided analytics tools for call centers into three categories, and we provide a few examples of each category throughout this article. One category comprises analytics software that incorporates middleware to gather data about calls, including lengths of calls, hold times and numbers of transfers. This data typically comes from phone switches but can also come from scheduling and call recording systems.

Another category includes tools that employ analytics to complement their primary function, which is to maintain a repository of interactions, whether by phone or on-line, between agents and customers. Instances of this type of analytics software are numerous, and we cite only a few examples in this article.

That's because the type of analytics software that offers the greatest promise for the near future encompasses efficiency within the larger context of performance. The third category in our overview of analytics tools refers to standalone software that lets you establish performance scorecards for individuals, for teams and for your overall call center.

We conclude this article by examining the evolution of analytics as a discipline from measuring efficiency based on data from phone switches to evaluating performance from the perspectives of your company and customers. We will follow up on this article later this year with our coverage of the newest performance management tools, as well as profiles of call centers that are broadening their performance goals beyond the realm of efficiency.

The Role of Middleware

The main criterion for analytics tools is that they present call center managers with information about how the agents they supervise manage their time while handling calls. In terms of delivering data, analytics tools should enable managers to schedule historical reports and receive real-time updates.

Another criterion for analytics tools is that they be able to gather data from diverse sources, including phone switches, call routing systems and interactive voice response (IVR) systems from multiple manufacturers and developers. Analytics tools should also save managers the work of aggregating data from different sources while allowing managers to maintain consistent definitions of metrics they report on.

One way to ensure the integrity of definitions of metrics amid a variety of data sources is to implement middleware. Symon Communications (Plano, TX), for instance, offers a new option with the latest version of its middleware software, Symon Enterprise Server (SES) 8.1. The new option is Series IV, which lets you indicate how you consolidate and aggregate data from different sources for the purpose of, for example, reporting on the efficiency of groups of agents or for devising scorecards that roll up various metrics into an index. (SES' other options enable you to select which data you report on, how you present it and where you present it, such as on electronic displays or within Web browsers.)

Like many middleware tools, SES allows you to send data to historical and real-time reports. When the data that feeds SES undergoes any updates, SES transmits only the data that's changed, rather than refreshing all the data SES collects, over your corporate network.

Symon provides a real-time reporting tool, Vista, and embeds a historical reporting tool, DataStore. DataStore is ideal if the data that feeds your reports resides within relational databases. If you prefer to look at data from multidimensional perspectives using third-party on-line analytical processing (OLAP) software, like those from Business Objects and Cognos, Symon offers additional software, DataMart, to feed your OLAP tools.

Inova Solutions' (Charlottesville, VA) middleware tool, Inova LightLink Enterprise, works with another tool from Inova, Data Scribe, that lets you format and organize data that you typically include within reports. In addition, Inova's Performance Tracker enables you to view real-time and historical data in the form of a dashboard. You choose the graphical format, such as a chart, graph, table or gauge, in which you present the data.

Inova provides these components as part of the latest version of its Performance Optimization Suite, which also comprises tools for disseminating data throughout your call center, as well as presenting the data on your colleagues' computers, on electronic displays and within Web browsers.

Most analytics tools for call centers that incorporate middleware enable you to define thresholds for individual metrics, or combinations of metrics. For example, the latest component of Centergistic Solutions' (Orange, CA) analytics suite, AgentView Enterprise Performance Management, lets you define thresholds for specific groups of agents or centers. This component, version 4.0 of Centergistics' AgentView Enterprise PowerUser, presents you with a dashboard from which you can select and view combinations of metrics and the scope of the metrics, whether they encompass centers, groups or individuals.

Keep in mind that although analytics tools alert you as to whether you're close to exceeding or have exceeded your thresholds, most analytics tools don't automatically start or stop procedures, like recording calls, related to these thresholds. For analytics tools to get past this limitation, the key component may very well turn out to be middleware, which can recognize a wide range of circumstances that require specific actions. In the near future, as several executives from among the vendors we interviewed have suggested, it's likely that analytics tools will be able to trigger certain processes, like adjusting call routing rules, in response to conditions you specify.

Report on Calls or Develop Agents? You Decide!

To broaden your options for measuring your call center's efficiency, you don't necessarily require middleware, nor do you need a CRM suite.

If you've implemented software to consolidate records of your company's conversations and on-line correspondence with customers, you may be able to apply the newly enhanced analytics capabilities that some of these communication tracking tools offer.

Among these tools is version 2.3 of Genticity's (Charlottetown, Prince Edward Island, Canada) Customer1, which now lets you break out metrics for different groups, as well as present dashboards of key performance indicators to agents.

For its clients that use its software to view information about communication with customers, Genesys (Daly City, CA) offers an optional component, Info Mart, that lets you create multidimensional reports from which you can drill down on data you've gathered about customers and agents. To use Info Mart, you also need to have third-party OLAP software tools, like those from Business Objects, Cognos and Hyperion, among others.

Astute Solutions (Columbus, OH) is enhancing the reporting acumen of its flagship software, PowerCenter, by enabling you to switch easily between normalized reports — which describe metrics in the context of others, like the number of support requests per sale — and non-normalized reports, which typically track individual metrics over time. PowerCenter also allows you to import and export data you report on from and to OLAP software.

What's becoming increasingly typical of analytics tools is that they are available standalone. That will be the case with Taske Technology's (Ottawa, Ontario, Canada) Call Visualizer, which lets you report on and color-code calls that you may need to look into further, such as calls that entailed transfers to multiple agents, or calls with unusually long hold times.

Taske currently provides Call Visualizer as a component of myTaske Reporter, which lets you view reports from Java-enabled Web browsers. Taske plans to offer Call Visualizer as a standalone product by the second quarter of this year. Call Visualizer will also let you view call-related statistics within Java-enabled Web browsers. As with myTaske Reporter, phone systems for small call centers will serve as primary sources of these statistics; these systems will include those from Avaya, Inter-Tel, Iwatsu, Mitel and Toshiba. Also like myTaske Reporter, Call Visualizer will be able to pull data about calls from workforce management systems from Blue Pumpkin, IEX and Left Bank Solutions, as well as Computer Telephony Solutions' call recording system.

Aspect Communications' (San Jose, CA) Aspect Data Mart similarly lets you gather and report on call-related data from phone switches from different manufacturers. In addition to switches from Aspect, they include those from Nortel and Avaya. Data Mart also lets you report on e-mail correspondence if you use e-mail routing software from eGain or Kana.

The most significant development with analytics tools is that their purview continues to widen. Last fall, Witness Systems (Roswell, GA) introduced standalone reporting software that gives you the option of drilling down on details of calls, including recordings of them, but doesn't require you to purchase Witness' call monitoring system to use this software. Witness' decision to provide this reporting software as a standalone component, especially in light of the company's acquisition of Blue Pumpkin, with its roster of scheduling and performance management tools, suggests a new direction for analytics.

We've remarked that some analytics software for call centers allows you to share data with third-party OLAP software so that you can view the data across multiple dimensions. That's true of Witness' reporting software, and the other tools we mention below, but in the near term, the more important requirement of analytics tools will be that they present you with a clearer understanding of your call center's performance.

Syntora's (Toronto, Ontario, Canada) Agentivity Analytics, for example, lets you create scorecards for agents, groups of agents or your entire center. The scorecards combine call-related statistics from Aspect's, Avaya's and Nortel's phone switches, as well as from Aspect's workforce management software, Avaya's predictive dialing system and Genesys' middleware. As with many analytics tools, Agentivity Analytics enables you to view real-time and historical data from within Web browsers; the software also allows you to build reports that factor in data related to calls, such as revenue or customer satisfaction scores.

Among Syntora's recent enhancements to Agentivity Analytics is a new charting module that lets you track, through graphical reports, trends concerning individual data elements within your scorecards. Since Agentivity Analytics can work with third-party computer telephony interfaces, many of which are now embedded in call routing systems and link directly to information about customers, Syntora's software enables you to build reports, for example, that track how efficiently agents assist customers who do the most business with your company.

Analytics software should allow you to combine and make use of a variety of metrics to gauge how your call center is performing. Besides tracking agents' efficiency, AIM technology's (Menlo Park, CA) AIMcall lets you link additional performance indicators for agents, such as feedback from customers, to agents' incentive plans. AIMcall also enables you to share performance statistics with Aspect's workforce management software and with Knowlagent's training tools.

From Analytics to Performance

We've outlined the capabilities of various analytics tools; it's equally helpful to acknowledge their limitations.

Analytics tools can consolidate and organize data in a format that enables you to spot trends. These tools do not identify root causes for you; at best they point up correlations among the data you choose to observe.

Because analytics tools for call centers grew out of reporting packages included with most phone systems, they are most adept at pulling together statistics concerning calls. As we noted earlier, if you focus on average speeds of answer, average talk times and average hold times, then you're strictly measuring performance from the standpoint of efficiency.

To answer the question of how useful these measures are in evaluating your call center's performance, let's draw an analogy between statistics in call centers and those used in baseball.

Like home runs in baseball, measures of efficiency are, within call centers, the most easily visible indicators of performance. Fans deem baseball players productive when they hit lots of home runs. But if these same players tend to strike out with runners on base, then their propensity to hit balls long distances becomes a liability.

Just as home runs alone don't win ballgames, efficiency alone doesn't win customers. When a ballplayer out in the field drops a ball, he has, in that moment, potentially cost his team a victory. If the player consistently errs in the field, his inability to perform in one part of his game undermines not only his own efforts, but also those of his teammates.

Similarly, when an agent communicates rudely or inaccurately with customers for the sake of handling a certain number of calls during the day, the agent renders customers — and therefore the call center — unproductive. More than detracting from his or her efficiency as an individual, the agent undermines the call center's ability to assist customers. In call centers, as in baseball, you can't be productive if you don't contribute to your team.

For any group of individuals — in a call center or a ballpark — to be successful, a manager must be able to account for a team's collective performance. Analytics provide a good start, especially in presenting information to call center managers. In terms of the day-to-day work of running a call center, there is value in having access to metrics, like schedule adherence and hold times, not only on a real-time basis, but also in being able to view them side-by-side.

"The way they want to consume that data has changed," Henry Lach, CEO of Syntora, says of call center managers. "We've seen demand for greater consolidation of information."

Since call centers are time-sensitive environments, those who run call centers often manage in response to exceptions. They look out for data that indicates problems that require their most immediate attention, like overly long hold times or agents who are not available when they should be.

There is some justification for this tendency. Lack of schedule adherence among agents, for instance, can result in long hold times for callers. When managers don't have sufficient data to track agents' efficiency, the outcome, especially among less patient callers, could be decisions to place calls — and their business — elsewhere. Time is money in call centers, and any information that can mitigate the risk of losing both is valuable. As Bill Howe, vice president of sales with Inova Solutions, says: "The good news can always wait."

What are the flaws of the management-by-exception approach? When call center managers focus exclusively on metrics related to short-term efficiency, they tend to share these metrics among their colleagues outside the call center. Problem is, the numbers that call center managers rely on to do their jobs aren't the numbers that their colleagues in other areas of their companies care about.

To convey a call center's value to a company and its customers, a call center manager has to describe the center's performance in terms that are recognizable to colleagues in other parts of the company. That's especially necessary, acknowledges AIM technology's CEO, Tony Hayward, since "the call center now affects every department in an organization."

At a time when every department needs to articulate its value, and prove it's not a drain on resources, call center managers have to be careful about what metrics they disseminate.

"There are very few call centers that can exist because they have to," says Doug Rippy, vice president of research and development with Astute Solutions.

The limitation of managing in response to exceptions is that it reinforces a reactive approach to running call centers. If your method of planning is to focus on quantities that are easiest to measure, like lengths of calls, and you don't consider why some calls are longer than others, then the simplest approach is to treat calls that exceed a specific duration as bad. When you manage reactively, says Tom Lockwood, senior product marketing manager with Aspect, "you tend to set some arbitrary target."

If you're a call center manager who looks for problems after the fact instead of seeking opportunities to serve a customer better, warns Lockwood, "potentially, you lose some valuable interactions with the customer." Oscar Alban, principal market consultant with Witness, concurs. He says that call centers that perpetually operate in triage mode "lose sight of opportunity." When the aim of your call center is to meet numbers, then you risk devaluing not only agents' efforts to assist customers, but your company's value to customers as well.

When call centers focus on continuous improvement, they no longer need to retreat into the safety of averages. Nor do they allow the risks of communicating with customers to overshadow the rewards.

The good news is that as companies come to recognize the effect of customer service on loyalty, they are striving to bring best practices in management to their call centers. Referring to one of the key concepts of Motorola's performance improvement methodology, Six Sigma (which we describe in another article in this issue), Alban asserts that "there is nowhere where 'Voice of the Customer' is more prevalent than in call centers."

For the voices of your customers to be heard, it's essential that you take the lead in articulating your call center's impact on your company's performance. To contribute to your own growth, and that of your organization, it's not enough to abide by averages, which only reinforce what your call center has done in the past. Expanding beyond analytics, performance management presents a framework for your call center's future.

As Jean-Marc Mezin, senior manager of product marketing with Genesys, says, "Tomorrow is more important than what happened today or yesterday."

---