Significant advances in technology make the QM process more effective.
Dictaphone, a company founded by Alexander Graham Bell and since absorbed by Nuance was the pioneer of call loggers. Other early players were Racal Recorders, ASC telecom, Lanier and Eyretel, among others. These devices were primarily used for public safety and liability management issues purposes. When deployed for quality monitoring calls would have to be selected from banks of audio tapes or supervisors would have to sit next to agents or remotely listen patiently for a coachable call. It was not until the late 1990s before a new generation of innovators introduced software that selectively retrieved calls for remote monitoring evaluation at the convenience of supervisors.
Limitations of Quality Management Practices
Quality monitoring software has continued to improve and today we estimate that between 60% and 70% of all contact centers and 95% of contact centers with over 200 agents deploy recording and quality monitoring software. That said, there are still many factors that limit the effectiveness of current quality management practices.
Evaluations are not aligned with corporate objectives. Using the same call selection, evaluation, and coaching techniques that have been used in the past provides a valuable historical record trends. However, most organizations do not exist in static economic and competitive environments. KPIs, soft skills and coaching methods need to be adjusted to meet current priorities.
Evaluations are not accurate. Even the best intentioned and most efficiently managed contact centers rarely review more than one in 200 or .05% of agent interactions. This is far too few to be statistically representative of the thousand or more interactions the typical agent performs each month. It is important to draw a distinction between performance reviews and quality monitoring reviews. The two are very different; the former conducted once or twice a year and used largely to assess the continued role of that individual within the organization. The latter has the very narrow focus of maintaining or improving interaction quality. Acceptance of this notion means that call selection should take a laser like focus on interactions that exhibit certain characteristics or encompass certain topics.
Agent resistance. Turnover is an unfortunate reality within contact centers. About one in five new employees will learn about the practice of quality monitoring for the first time. There will be sensitivity to the “Big Brother” view of someone listening to their conversations. It is very important that management explain the need for call monitoring and enlist the support of agents in constructing the forms and driving coaching and training actions.
Drain on supervisor time. A supervisor responsible for 15 agents who need be monitored five times a month will conduct 75 evaluations each month. If each takes 30 minutes to complete and review with each agent, the process would require one full week of each supervisor’s time. Every day brings new challenges that need immediate attention. This inevitably leads to postponing the evaluations toward the end of the month. Coaching, if it occurs at all, is conducted long after the coachable event. The drain on supervisor time can be reduced by encouraging peer and self reviews, both of which tend to work very well and help build agent morale. Management should also take a close look at the evaluation form itself. There should never be more than 20 questions or rating scales and if possible this number should be reduced.
Perceived unfair evaluations. Monthly reviews consist of both objective metrics and subjective judgments. Agents are expected to perform well on soft skills such as courtesy, listening, empathy, problem-solving and the like. Different supervisors or QM analysts will have different criteria on which they base these judgments. Some agents will perceive favoritism on the part supervisors. Well-organized calibration sessions, ideally including a small number of agents, will help build consistency in these ratings.
Technology Remedies
Significant advances, too numerous to outline this article, make the quality management process more effective.
Intelligent search and retrieval. Most vendors today attach a wealth of metadata to each recorded call which allows drill down to very specific interactions. For example, VPI EMPOWER integrates with CRM or Helpdesk customer servicing applications to attach data such as customer ID, ticket # and call disposition/reason to help. By searching on customer identifiers such as these analysts can help to understand the causes of repeat calls.
Speech analytics. Speech technology is capable of organizing voice interactions by category. The more sophisticated systems can self-create categories. Less costly word- and phrase-spotting techniques require a little help in creating the categories. Selecting calls by category provides a very valuable search tool, highlighting types of interactions that require more coaching assistance. Some systems even have a means of detecting changes in emotion. Emotional intensity is presented in graphical form which tells the analyst exactly when to interrogate the recording
Integration with voice of the customer data. While traditional KPIs remain essential barometers of performance, an increasing number of contact centers are also directly measuring the pulse of the customer. Customer satisfaction surveys are rapidly becoming standard in contact centers and the results are linked to agent interactions. When combined with speech analytics it is possible to obtain a clear view of what factors drive customer satisfaction.
Flexible evaluation form builders. Authoring tools allow management to maintain different evaluation forms. This is important because one size does not fit all. Some agents specialize in certain tasks, some are multilingual, some are experienced and others may still be learning their new position. Perhaps more importantly, the authoring tools allow for quick adjustments based on changing priorities. VPI uses a “rules engine” to help decide which interactions need to be reviewed and automatically assigns them to the responsible supervisor or quality analyst. It can recognize the type of interaction that is setting off the rule and automatically suggest the most appropriate evaluation form to be used for the inspection. (See Image 1.)
Where Does Quality Management Go From Here?
Looking to the future we can anticipate advances at least four areas;
Multichannel. The “customer journey” may begin at a website or IVR and then travel over other voice and data channels until the issue is resolved. The quality management process and enabling technology needs to provide meaningful metrics for each leg of this journey.
Social media. Many organizations invite commentary on their Facebook and other social media sites. There are continuing questions about whether a company should respond to these inputs and if so how the responses should be evaluated. Leading companies such as Verint Systems are working on solutions for interrogating and evaluating social media interactions. Verint Text Analytics can quickly sort through thousands of unstructured data points including social media comments to surface enterprise customer service issues that demand immediate attention, as well as those that are starting to bubble beneath the surface.
Video communications. As video continues to grow as a significant channel of B-to-C and B-to-B communications so also does the need to apply the same level of quality management now reserved for voice and data communications. ZOOM International, in collaboration with Cisco, is ahead of the curve by providing the tools you need to intelligently retrieve stored video and measure quality against desired standards
Automated quality management. Innovative vendors are using automation to reduce demands on supervisor time and help deliver more accurate and more consistent appraisals. CallMiner, a leader in speech technology, has developed software that automates much of the quality management process. The system monitors all of agents’ voice interactions, not just a sample, and automatically rates the aggregate interactions on soft skills using consensus criteria. (See Image 2.) This frees more time to personally review calls that exhibit anomalies and interactions handled by recent hires. Automation does not replace human evaluations, but it does allow them to spend more coaching time addressing the issues uncovered from the automated scoring process.
We expect innovative vendors to continue to find ways to streamline the quality monitoring process and to launch integrated solutions that can identify pain points and root causes that encumber the consumer from enjoying a smooth journey to the enterprise regardless of paths taken.