By : Jim Beuoy, Director of Quality Assurance and Corporate Compliance and Dan Werner, National Sales Director
Managing performance in the call center calls upon a number of different inputs. We basically see these influencing factors as three distinct categories: 1) people 2) process 3) technology. There are many technologies available to assist the QA Department and all areas of a company to better manage their agents. We could spend pages and pages talking about all of these, but for the purpose of this article, we’re going to focus strictly on the human element. More specifically, we’re going to discuss alternative philosophies and approaches to monitoring, coaching, and scoring agents.
Every organization struggles with the balancing act between production and quality. Folks from the quality side will often argue that doing things right and doing them right over and over will make the numbers take care of themselves. Folks from the financial side will see coaching and training as cutting into (payroll) hours that could be invested in revenue generation. There’s no question that coaching and training can be a significant expense, and if allowed to run unchecked can have a diminishing return on the investment. Vice versa, allowing production to run unchecked, producing (billable) hours for the sake of billing hours is not unlike the manufacturing philosophy of the 1950’s that failed to take into account the financial impact of the true cost of scrap, customer satisfaction, and a host of other bottom line influences. The fact is, not every one can afford a Lexus. Some budgets can only afford a Saturn. You’ll obviously want to build a very good Saturn, but the bottom line is that you have to balance agent development / improvement hours based on economic constraints of the company.
When companies first embark on a QA initiative, we often see a tendency to form a large group of monitors who are saddled with the responsibility of “managing” quality for the company. However, there’s a lot of baggage with this concept. One, it’s enormously expensive. As we previously noted, not every client can afford this service. Some companies have recently branded the QA function and sold it as an optional feature. The point is that the overhead of a large monitoring group can make you less competitive in the market place.
The second downside of a monitoring / QA department is that it can create animosities between departments. Agents can get the impression that “big brother” is always watching, just waiting for the opportunity to catch us doing something wrong. And if you think about it in the context of the 1950’s manufacturing, this approach shifts QA responsibility from an in-line process to a after-process, detection and correction method. We’re of the opinion that QA is everyone’s responsibility and needs to stay in-line as much as possible. It’s critical for the agents to perceive QA initiatives as something for them: an effort to help them with the skilled tools of their trade with an objective of continuous improvement.
We see quality assurance as having 3 key members. Those key members are the Agent, the Supervisor, and the Quality Assurance Team. Below are a few key points to help you better measure your call center’s performance:
1. Role of Supervisor vs. Quality Assurance Department (QA)—The Supervisors and QA work together to maximize performance of their agents, but each have separate roles and responsibilities:
Supervisor— Obviously, the Supervisor is charged with meeting predetermined objectives and KPI’s. In our environment, the Supervisor also owns the overall quality of their team, identifies opportunities to improve each agent, and modify agent behavior through coaching.
QA—The QA team is not the police. Their job is to monitor for compliance, establish company standards, provide a calibration function, and to identify company-wide trends and areas for improvement. They should be positioned as an asset to Service Delivery / Production. Their monitoring forms don’t replace the evaluations done by the Supervisor. They only supplement the Supervisor’s evaluation of their own team members.
It’s our philosophy that leaving QA responsibility mostly with the Supervisor and the agent establishes bonding with the team and enhances behavior modification for continuous improvement.
2. Why do we Monitor Calls—We monitor calls to ensure that agents deliver accurate information, all of the desired information, in a professional manner, in a way that will produce satisfied customers, and in a way that leaves a favorable impression on the caller. We also monitor for script improvement, to gather marketing intelligence, and to obtain a company quality score. That brings to mind an important person in the mix that we haven’t discussed so far. That’s the Project Manager. The Project Manager, being the person who answers to the client (in a Service Bureau setting) has final responsibility for the project’s success or failure. It clearly behooves the Project Manger to be part of the quality team. The Project Manager should also monitor agents but from a different perspective from the Supervisor and (QA) monitors. The Project Manager should be listing for questions for which there were no answers provided in the training kit / on-line resources. He/she should be listening for objections for which no answer has been provided to the agents. In short, the Project Manager monitors to determine if production was given the proper tools. The production supervisor and monitors listen to determine if the agents are using those tools appropriately. Of course supervisors and monitors can be a invaluable feedback to the Project Managers in terms of things they are running into, for which they we were not prepared. The point is that multiple people on the quality team have monitoring responsibilities, however, their specific focus may differ.
3. Preparation for Monitoring Calls—Preparation is key in order to give you a basis for effective improvement of agent performance. One would never simply monitor a call and then go coach on that call. Besides the obvious flaw of one call not being a valid sample size, this could lead to coaching on an aberration / an isolated agent behavior. Instead, we need to first review past coaching history and performance statistics. Performance statistics are indicators of behaviors. It would take a whole article to delve into what behaviors each KPI indicates, so we won’t go into that detail here. We simply need to note that you draw assumptions about agent work habits and/or skill sets from productivity reports before you monitor and coach. Manage what you expect and measure what you want to manage. Monitoring should be attune to KPI indicators so that you end up coaching on the behaviors that have the largest impact on agent efficiency and effectiveness - not some isolated incidence. A centralized repository provides us with all of the detailed analytics showing agent performance and team performance on each project so we have a baseline measurement for the monitoring session. The coaching logs are also reviewed to see what the goal was from the last coaching session. We don’t want to go onto an endless number of unaccomplished goals. We build upon the success of reaching previous goals in order to tackle more complex or difficult skills.
4. How to score calls—We often see scoring systems with broad ranges of weighting systems. A little later in this article, we’ll talk about why and how to conduct calibration, but suffice it to say for now that it’s important to come up with the same score, regardless of who does the scoring. The score shouldn’t be determined by who’s doing the scoring. It should be a function of what the agent does or doesn’t do, right? If a scoring system requires the evaluator to select a score from 1-10, it’s very likely that you will have a significant deviance in your respective scores. Wide ranges make it difficult to determine, let alone defend the difference between, say, a 5 and a 7. There isn’t an industry “standard” per se, but there seems to be a consensus that a more narrow range leads to greater consistency across scorers. We strongly recommend an even number, such as 4 or 6 in order to eliminate the tendency of scores to migrate to the middle. Force them to make a decision of right or left of the median. We also believe that the score should not be a number as much as it should be a definition. (Best practices) top scores for example, often have a definition of: Excellent, Always Exhibits, Exceeds, or WOW! On the opposite end of the spectrum, we strongly encourage staying away from negative ratings such as Doesn’t Exhibit, Doesn’t Meet Requirements, or Couldn’t deliver a script in a bucket. (Just kidding about the last score – we haven’t really seen that one, but you get the idea.) You want to make sure that the coaching experience is a positive one; one of continuous improvement. You’ll want your low-end scores to support that with more positive definitions such as “Opportunity Area”. If you’re using an electronic scoring system, you should easily be able to change the text rating to a numerical score programmatically.
We’ve also seen scoring systems with categories such as greeting, professionalism, and product knowledge. As with a wide weighting system, the challenge with broad categories like product knowledge and professionalism is: How would each scorer define them? Just as we suggested that best practices is to define score weights, we propose that scoring systems define product knowledge and professionalism. We see them as two distinct qualifications. An agent could, for example, deliver all the necessary information in the script but be monotone and boring or perhaps not be empathetic. In either case, the agent is not likely to be as effective as he or she could be. And vice versa; an agent could sound interesting and interested (in the customer) but leave out legally required disclosure. In either case, the rating system should indicate a failed experience. In our scoring system, we actually give the agent two different scores: one for what they said and one for how they said it. This is done by measuring the content of the call and the quality of the presentation in order to create a combined quality score that gives you a true indication.
5. Measuring the Content of the Call---Take your script and break it down into meaningful components. Let’s use an inbound call as an example. Did the agent say or did the agent not say: “Thank you for calling”? Did the agent say or did the agent not say the company name? (…assuming that you want them to say the company name. There may be scenarios where the agent is expected to use a more generic name such as “sales department” because agents handle calls from multiple lead sources, campaigns or whatever.) Did the agent give his/her name or not? These should be absolutes. They either use the script and the language that you want them to use or they don’t. Absolute, presence or absences scoring is called baseline scoring. Our system uses baseline scoring for content of the call.
When it comes to measuring product knowledge, you’ll want to go beyond scoring just the meaningful components of the script. You’ll also want to have questions such as: If the agent got an objection, did he/she use the approved response? And, If the agent received a question, did he/she give the correct answer? Again, if you don’t define how you want to measure product knowledge (perhaps in terms such as these) you’ll end up with gaps between scoring and questions about the validity of your (inconsistent) scores.
6. Measuring Quality of Presentation—This is our definition of professionalism. We decided that professionalism needs to be defined in terms of certain attributes that the agent demonstrates. Our scoring system includes:
Volume—Extent to which the agent speaks in a volume that is neither overbearing nor so soft that the prospect has to strain to hear.
Pacing—Extent to which the agent adjusts rate of speech to match the style of the prospect.
Grammar—The extent to which the agent uses appropriate words and syntax
Pronunciation—The extent to which the agent correctly enunciates each word with emphasis on the correct syllable
Tone—The extent to which the agent utilizes appropriate tone, inflection, empathy / enthusiasm for the occasion
Listening—The extent to which the agent allows the prospect to speak and demonstrates an understanding of the consumers concerns.
Obviously these criteria cannot be measured in absolute terms. “The extent to which” or “The degree to which” clearly indicates that a scoring range must come into play. This is called finesse scoring. As previously noted, we suggest an even number of ratings, with clear definitions for each rating.
7. Providing Feedback—Positive feedback should be your goal while also constructively pointing out areas needing improvement. The agent should want to receive this feedback in order to continually improve. Common pitfalls include monitoring the wrong agents too often, nitpicking, talking too much, robbing the agent of their dignity, and focusing only on errors and mistakes. Positive language should be used at all times and use the 3rd person when possible. We like providing feedback to agents in a private setting, where potential embarrassment by peers within ear shot can’t distract from the learning experience.
Best practices indicate that agents should provide their own feedback! If an agent hears him/herself, and indicates: “I never knew I sounded that boring!” or “ew, I should have used the No Money objection response right there!” the coaching session is much more likely to be perceived as a positive experience than if the Supervisor tells the agent everything they did well and could have done better. Self-discovery is huge in other ways. The Supervisor is no longer the bad guy, delivering all the feedback. Ask yourself: Would you rather be asked or be told? Just about everyone would prefer be asked. You’ll find that agents are typically harder on themselves than any supervisor or monitor. Furthermore, if the agent indicates that they have an opportunity area and that they want to succeed in improving in that area, behavior modification is MUCH more likely to happen. When it comes to changing behaviors, it doesn’t matter what the Supervisor thinks or what the Monitor thinks. It only matters what the agent thinks.
8. Calibration—This process will help you to maintain fairness, objectivity, and consistency in scoring your calls. The calibration sessions should include the Program Manager, the Supervisor(s), the QA Team, the client, and the Agents. You can tell that your calibration is working when there is little or no gap between how you score the calls, how your team scores the calls, and how the end client scores the calls.
For further information, please contact Dan Werner at 866-671-0778 or dwerner@oksameridial.com