|
Not sure that it helps but a good analytics is a key for good performance. Check this greate article found elsewhere in internet.
New Methods to Measure Performance
With the emergence of speech and predictive analytics, call centers have many more ways to evaluate their communication with customers.
By Joe Fleischer
02/01/2007, 5:00 AM ET
What do we mean by call center analytics? Before we start talking about analytics software for call centers, we have to define analytics to enable us to refer to metrics that are directly relevant to call centers, and to place these metrics in the larger context of a call center's performance.
During the last few years, the definition of analytics has evolved from emphasizing measures of efficiency to combining indicators of efficiency and quality. Today's call center analytics tools don't necessarily restrict themselves to counting calls and indicating how long they are. These tools accommodate not only quantitative data, such as average handle time, but also qualitative data, such as scores supervisors assign to the agents they evaluate.
The first step in making sense of call center analytics is to come up with a set of metrics that are within the immediate scope of call centers. In many centers, a successful call is one where an agent fulfills a customer's need and, at the same time, generates or retains revenue from the customer. The challenge many call centers create for themselves is that they try to identify one single metric that encompasses all the characteristics of a successful call. As we explain below, such a metric doesn't exist, which is why call center analytics refers to an index of metrics.
What should this index consist of? Let's consider average handle time, which is how long it takes for an agent to communicate and to document the results of the interaction with a customer. The more conversations that occur between agents and customers, the more opportunities agents have to generate or retain revenue, so, in theory, it seems logical for a call center to use average handle time as an indicator of performance. In practice, average handle time, along with the hourly cost of employing the agent, tell you how much money you pay the agent per call. But this metric reveals little about whether the agent knows how to communicate with customers.
Similarly, other metrics, like the percentage of time the agent adheres to his or her schedule, and the percentage of time the agent is handling calls from customers, tell you how busy, but not necessarily how productive, the agent is.
In a setting where the goal of a call, for instance, is to resolve a request for support, the call center needs to define productivity from the customer's perspective, rather than from the agent's point of view alone. That's because, by definition, a resolution means that the agent fulfills the request to the customer's satisfaction. Even when we define productivity based on revenue from a customer who calls to place an order, we have to consider if it's possible for some agents to generate more revenue than other agents. If so, then we might want to figure out how to replicate methods, such as upselling and cross-selling, that the most productive agents employ to bring in additional revenue.
Some call centers include internal evaluations of recorded conversations between customers and agents among the metrics they use to gauge an agent's productivity. A call monitoring form usually contains a checklist of behaviors that are easy to observe and document while listening to a recording of a call. Some behaviors, like asking customers at the start of a call if they're members of a loyalty program, contribute directly to a company's long-term revenue, and therefore reflect agents' productivity, so they often appear on agents' evaluation forms.
Internal evaluations are subjective and they only refer to the perceptions of supervisors or quality assurance staff within call centers. But they do indicate if agents are communicating in a way that the company believes can accomplish the goals of calls as efficiently and consistently as possible. External evaluations, such as customer surveys, provide a valuable reality check; they enable companies to find out what effect, if any, their communication with customers has on whether customers want to do business with them.
Using the examples above, we've summarized the types of metrics that would be useful for an index that accounts for various aspects of a call center agent's productivity. We must now distinguish between types of analytics that call centers track in most circumstances, and those that call centers pay attention to concerning certain types of calls or as part of other efforts.
Speech analytics, for instance, can be useful for streamlining the process of monitoring agents. That's literally the case, for example, when agents are under a legal obligation or corporate mandate to follow scripts during any portions of conversations with customers. More generally, if we believe that the result of a call hinges on whether customers or agents say certain words or phrases, then speech analytics can help speed up searches for calls that contain language that we associate with specific outcomes. But it's important to remember that speech analytics alone reveals nothing about how agents do their jobs. Speech analytics can help you evaluate agents, but doesn't replace evaluation.
Predictive analytics, as the name of this category suggests, enables you to calculate the likelihood that customers will pay their bills, respond to certain promotions or simply remain customers. A variety of corporate departments, such as marketing, rely on predictive analytics software regardless of whether these departments collaborate with call centers. Even when the implementation of predictive analytics software entails gathering information from call centers, the decision to deploy this type of software is one that involves the entire company, and doesn't necessarily occur within the specific realm of call centers. Still, there's a good chance that this year, we'll start to see more call centers use predictive analytics alongside other types of analytics tools.
Step Two -- The Context of Performance
Now that we've established the types of call center metrics that belong in an index, and now that we've outlined the categories of analytics we refer to, we're ready for the second step in our efforts to understand analytics, which is to place these metrics in the larger context of performance.
From a tactical perspective, analytics tools allow call center managers and certain colleagues, such as business analysts, to view historical and real-time information about metrics concerning the agents and the operations managers are responsible for. Analytics tools indicate trends and correlations over time, whether these trends refer to individual metrics or indices that tie together multiple metrics.
As a category of software and as an overall discipline, analytics is a subset of the more strategic realm of performance management. Analytics is descriptive in conveying how metrics vary not only over time, but also among individual agents and groups of agents. Performance management, by contrast, is prescriptive; it suggests actions that call centers need to take to change or sustain the trends that analytics tools describe.
"Analytics, for example, might focus on the service level for the previous half hour or an agent's real-time adherence to schedule," says Brett Williams, senior manager of performance optimization product management with Aspect Software (Westford, MA). "Performance management might focus on improving customer satisfaction for the year by aligning agents' operational activities to support the customer satisfaction improvement initiative."
Matt Schwabel, marketing director with Inova Solutions (Charlottesville, VA), believes that to understand analytics, call center managers first have to understand performance management, which he defines "as a process of continually measuring, analyzing and modifying [agent] activity in order to achieve established business goals." He explains that analytics is "an aspect of performance management," and he defines analytics "as a method of looking at pertinent data in a way that reveals trends and insights that have business impact."
Schwabel adds that analytics isn't merely about displaying data, but rather about representing data in a way that's meaningful to call center managers and to the overall business the call center is part of. He says that it's "important that this data is not simply presented in a raw form, but that the tools offer the ability to combine and manipulate the data into relevant metrics."
Tony Hayward, CEO of San Francisco-based AIM Technology, suggests that although analytics is a subset of performance management, the two are intertwined. "From our perspective, analytics are a key component of performance management, which outlines the bigger issue of optimizing agent performance to meet the business and operational goals," he says. "Analytics in the call center is a business and operational technology tool that allows everyone from team leaders to analysts to executives to dig deep into the operational systems to acquire a higher depth of knowledge of information regarding individual, team or center level performance."
"Often the data that provides that level of insight must be accumulated and correlated across different operational systems," explains Hayward. "The value in analytics is correlating the data from those systems together in a fashion that uncovers previously unknown knowledge about performance."
Dan Derin, president of U.S. operations with Genticity (Charlottetown, Prince Edward Island, Canada), concurs. "Analytics [enables] performance management and performance management drives the definition of analytics," he says. "The main objective of performance management is to align business process goals with overall organizational goals. These goals include profitability, market competitiveness and customer loyalty. When these goals are stated in clearly measurable terms, analytics become the vehicle for achieving them."
Trends and Tools
According to executives from vendors that develop analytics software tools, one of the biggest challenges call centers face is ensuring that information about interactions with customers is easily accessible, yet meaningful to those who view this information.
"Our customers' biggest pain tends to be getting all their data into one place where they can report on it in a consolidated manner," says Chris Crosby, president and CEO of Chicago-based Latigent. "By creating predefined templates that can be filtered for individuals' relevant scope of responsibility, you can put the power of opportunity identification in the hands of even the least technical and analytical individuals in an organization."
Another important attribute of call center analytics tools is that they display metrics that correspond to a supervisor's area of responsibility. Hayward points out that version 6 of AIM Technology's AIMPerformance "pre-determines the required analytics views of information for a specific business user or business situation."
Call center analytics tools also let companies shed light on the connection, even if it's simply a correlation, between a call center and other areas of a company.
"Analytics at its most generic is a process of interacting with data in order to tease out results you might not have revealed with a report," says Tom Lockwood, senior product marketing manager with Genesys (Daly City, CA). "Rather than starting with a hypothesis and then writing a query to validate that hypothesis, analytics presents you with a data-rich environment for interacting with data, allowing you to uncover facts relevant to the efficiency of your operations."
Susan Peter, vice president of marketing with HardMetrics (Doylestown, PA), cites the example of a multibillion dollar cable company that uses her company's software "to analyze its Cumulative Leakage Index (CLI), which measures how much a cable plant leaks into the environment."
"The proactive CLI monitoring allows the cable giant to adhere to FCC regulations and identify faults quickly," Peter explains. "Additionally, HardMetrics has allowed the cable company to seamlessly map between the CLI engineering system back into call center systems, thereby enabling real-time communication and collaboration between the engineering division and the call center. The engineering team is now able to flag customer service representatives on problems even before a customer calls in a complaint. By doing so, front line agents have access to better information, which helps them resolve a customer problem on the first call without having to deploy expensive field service agents to the customer premise."
A number of vendors are broadening the scope of their analytics tools this year to make it easy for different areas within a business to gather and share information that call centers possess about customers. Genticity's Derin says that Genticity plans to "release a new business intelligence framework to support call center analytics that is highly configurable to adjust to a variety of operating models." He explains that "users will have the option to subscribe to reports deliverable in the media of their choice."
As some vendors acknowledge, some call centers use analytics specifically to report on their own operations. In these centers, call center managers primarily use analytics to look up historical reports, and to view graphs of real-time metrics, or indices of metrics, within onscreen dashboards.
"I think reporting systems that make sense and have value across the organization have always been on senior management's wish list," says Bob Brittan, director of product marketing with Plano, TX-based Symon Communications. But he also acknowledges that "going back to our legacy products and real-time reporting, [call] centers have seen that value and now want more sophisticated systems for added visibility and intelligence."
Yet even as the practice of analytics expands beyond its emphasis on efficiency, a number of call centers still place a priority on tracking costs per call. "Our latest new module provides communications cost management, because cost will forever be an issue in [call] centers," says Susan Saldibar, vice president of sales and marketing with Centergistic Solutions (Orange, CA). "Managers want to know how economically transactions are being handled."
The Role of Other Types of Analytics
Several vendors that provide call monitoring systems as part of broader suites have sought to make the case that speech analytics is a key component of performance management.
"One of the newest and most interesting trends we are beginning to see is the combination of speech and performance analytics," says Daniel Ziv, vice president of customer interaction analytics with Verint Systems (Melville, NY). "In this scenario, speech analytics is used to tag calls by category based on what is said in the call, and then performance analytics establishes correlations between call category and other factors."
David Pennington, director of product management with Seattle-based Envision, concurs. "We are seeing stronger interest in speech analytics as a way to extract intelligence from recorded agent-customer interactions," he says.
Nancy Treaster, senior vice president of global marketing and product management with Witness Systems (Roswell, GA), points out that analytics tools, including speech analytics modules, establish a structure that call centers can use to categorize conversations with customers.
"Analytics is more ad hoc than performance management and may include unstructured data," says Treaster. "It helps organizations identify business challenges and customer issues based on data captured in the [call] center. One of the more recent best practices in evaluating performance includes gathering external customer feedback and comparing that to internal views of performance. And less-adopted but often-talked-about speech analytics products are being used to search for information such as customer issues or performance problems against large volumes of unstructured data."
In referring to speech analytics, Treaster maintains a clear distinction between analytics and performance. "While analytics allows for exploration of information, it does not measure the results against expectations like performance management," she explains. "Performance management, on the other hand, is more structured. It involves a series of metrics to ensure that the [call] center improves productivity and meets its departmental objectives. It then supports departmental objectives with key performance indicators that are associated with the goals. Performance management also offers automated data collection and reporting technology that frees up [call] center managers to work toward achieving goals."
Roger Woolley, senior vice president of marketing and chief marketing officer with etalk (Irving, TX), finds that etalk's "clients are realizing that unstructured information can be explored, and priorities are changing to include this data in the decision-making process."
"With the evolution of speech analytics and performance management, users are transforming the way information is evaluated and leveraged across the enterprise," he says.
Where do speech analytics tools come from? The parent of etalk, Autonomy, develops speech analytics tools; etalk employs these tools as part of its Qfiniti Explore module. Witness Systems, along with VPI (Camarillo, CA), are among performance management software vendors that partner with CallMiner (Fort Myers, FL) to offer speech analytics modules to their clients. Envision collaborates with San Francisco-based Utopy to furnish its clients with speech analytics tools.
Utopy in particular has branched out into yet another realm: predictive analytics. A number of other vendors are also touting predictive analytics software for call centers.
Tom Evans, a product manager with Austin, TX-based Austin Logistics, says that predictive analytics software tools gauge "which customers are retention risks, and which have a high propensity to accept cross-sell or upsell offers." Evans adds that predictive analytics tools are also applicable within outbound call centers to help them "determine the best time to call each customer to increase the probability of getting a right-party contact."
Referring to PredictiveCallCenter from Chicago-based SPSS, Chet Friedman, the company's director of product marketing, says that the software "integrates with existing call center systems to instantly determine which inbound callers are the best candidates for an upsell, cross-sell or retention offer, and which offer to make."
He adds that the software "automatically provides a recommendation or suggested action on the agent's screen, along with sales arguments and other information the agent needs to close the sale." In addition, says Friedman, "each recommendation balances the likelihood of customer acceptance with the potential value the offer holds for the company."
Predictive analytics is available not only as a product, but also as a service. For example, Assurant Solutions, an Atlanta-based outsourcer, offers its call center clients a predictive analytics service known as Targeted Solutions.
How can call center managers combine new approaches, including speech and predictive analytics, to gain a broader understanding of how agents perform? This is a question we will strive to answer throughout the year as we observe the ongoing evolution of key performance indicators and performance management in call centers.
--------------------------------------------------------------------------------
How Supervisors Improve Performance
Few people would disagree with the notion that to evaluate a call center accurately, and to improve the center's performance, supervisors need to do more than generate reports about the average amount of time agents spend assisting each customer.
But is this assertion correct? To find out, Merced Systems, a Redwood Shores, CA-based developer of performance management software, conducted an online survey last year. The survey received 129 responses from managers from 109 companies, whose call center operations varied in size from 250 or fewer agents to more than 5,000 agents. (None of these companies were clients of Merced.)
One of the most revealing results of the survey is that 57% of respondents say they don't believe that call center supervisors in their organizations devote enough time to coaching and developing agents. More specifically, 75% of respondents say that supervisors aren't effective at using data to determine priorities for how they coach and develop agents.
What makes these findings especially significant is that the survey notes a correlation between organizations where respondents say their coaching and development efforts are consistent among teams and call centers, and organizations where respondents say they have exceeded their goals during 2005 for revenue, customer satisfaction and internal quality scores.
Matt Katz, vice president of business consulting with Merced, shares the results of this study during a 60-minute Webcast called "How to Manage Performance Instead of Numbers," which you can view by clicking on "Webcast Archives" on callcentermagazine.com.
In response to an attendee's question during the Q&A portion of the Webcast, Katz summarizes the different ways that organizations typically determine agents' performance metrics.
"In the organizations we've worked with, we've seen some empowered reporting and analytics teams, where they're given the tools and the company's endorsement to mine the existing data and look for any correlations in their existing performance data if they have it," says Katz. "In other organizations, they talk to the line management that's had the experience in watching firsthand what measures the agents respond to. So we've really seen across the spectrum of both business and IT in terms of the responsibilities. It's clear, though, that when these metrics are tied to company and business goals, that they have the highest results or have the biggest impact."
During the Webcast, Mark Steinweg, general manager of Carlson Leisure Travel Services, describes the evolution of performance indicators for agents within his organization, which implemented Merced's software. "We started with the very traditional efficiency measures around things like average handle time; that's where we focused our initial effort," says Steinweg. "Along the way we also were able to put results of our quality assurance program into the reporting mechanism." He adds that his company's call centers now also incorporate among their performance indicators "revenue generating metrics such as upsell and cross-sell."
If we think of performance management as a category of software, then we're referring to a process of consolidating data that lends itself to automation. By contrast, when we think of performance management as a discipline, then we're referring to an approach to evaluation that requires human judgment. Yet, as Steinweg explains, by enabling a supervisor to spend more time with agents, and less time putting together reports about agents, the decision to automate the aggregation and analysis of data gives the supervisor the opportunity to focus on areas, like coaching, that have the greatest effect on a call center's performance.
"We've taken 30-40% of the time that they were spending with manual data gathering and freed that up to allow them to coach more; they are clearly spending more time with their teams," says Steinweg. "We've also been able to increase the span of control of the supervisors. Because we've taken a lot of those manual tasks off, they can manage a larger workgroup of agents."
--------------------------------------------------------------------------------
Analyze This
The following companies offer analytics tools for call centers.
AIM Technology 415-692-5580
Aspect Software 978-952-0200
Assurant Solutions 770-763-1000
Austin Logistics 512-328-8215
CallMiner 239-689-MINE
Centergistic Solutions 800-387-0264/714-935-9000
Envision 206-225-0800
etalk 888-258-1528/972-819-3100
Genesys 888-GENESYS/650-466-1100
Genticity 866-55-CUST1
HardMetrics 215-297-9738
HigherGround 818-591-3133
Inova Solutions 866-686-8774/434-817-8000
Latigent 866-LATIGENT/773-572-8709
Merced Systems 650-486-4000
Nice Systems 866-999-NICE/201-964-2600
SPSS 312-651-3000
Symon Communications 877-796-6634/972-578-8484
Telrex 425-827-6156
Texas Digital Systems 800-693-2628/979-693-9378
Utopy 866-44-UTOPY/415-621-5700
Verint Systems 631-962-9600
VPI 888-642-3774
Witness Systems 888-3-WITNESS/770-754-1900
|
Forum 
Favoured: 0
