Reimagine Your Quality Approach

According to a survey conducted by Execs In The Know (EITK) and COPC Inc. for the 2019 Corporate Edition of the Customer Experience Management Benchmark (CXMB) Series, only 48% of customer experience (CX) leaders said their quality program’s results were aligned with customer satisfaction and only 43% said definitively that their quality efforts drive improvements in customer satisfaction and issue resolution. This raises a question – why would you continue to invest so significantly in an activity that is not adding value to your organization?

In fact, one of our very own community members, Thomas Siebert., commented in an online discussion about this very topic. Thomas has worked with global brands and outsourcing providers so has experienced first-hand the effectiveness of quality programs from both perspectives:

“Most QA programs I’ve experienced were not effective at driving significant and sustained improvements and were actually set up to fail. This is largely because they were not designed or executed to create an internal quality standard, but were designed over time with extraordinarily complex questions, definitions, scoring, and overall execution. Additionally, most programs I have experienced were not directly tied to company policy and procedure, and seldom could one effectively discern product, process or policy level strengths and weaknesses. Most importantly, these programs were not set up to accurately capture customer pain points. The analytics were also often poor and dangerously conflated customer satisfaction and quality measurement.”

We suspect Siebert is not alone in these challenges. In this article we will explore ways you can re-imagine your quality program immediately to provide your organization with more actionable insights, leading to significant and sustained improvement. We will also provide food for thought about the future of quality as you plan for the long term.

History of ‘Call Centers’ and Quality Monitoring

To set the stage, let’s take a stroll down memory lane.

Graphic Source: EXL Service

It is of no surprise that call centers have dramatically evolved over the years, both in name and function. It was not that long ago that call centers were only handling customer interactions by phone (hence the name) and even some by mail and fax. Next, we saw the introduction of interactive voice response (IVR), email transactions, and then chat. In more recent years, we saw the explosion of multi-channel offerings such as social media, mobile applications, SMS/text, and now artificial intelligence (AI)/machine learning/chatbots. In the last few years, omnichannel became the focus with a shift toward greater personalization, more digital/self-help solutions, and advanced analytics.

Quality monitoring has been an integral function within the call center since the beginning and in many ways it has also evolved. However, some would argue it has lagged in approach, sophistication, and value.

We hear that many leaders find themselves in the position of justifying their current quality monitoring programs because it is perceived primarily as a cost of doing business, versus a value-add with a quantifiable return-on-investment (ROI). Who can blame them when, on paper, quality programs can be very costly when you consider:

  • Dedicated quality staff (if applicable)
  • Monitoring time (if quality monitoring is conducted by supervisors)
  • Quality monitoring technology (screen recording, scoring systems, reporting)
  • Coaching and feedback time (by quality staff or supervisors)
  • Calibration time

This can add up to hundreds of thousands of dollars and possibly more with global organizations.

Despite the high costs and significant time involved, many organizations experience these five common issues:

  1. Quality scores that do not align with reality – high quality scores but low customer satisfaction.
  2. Action plans are developed based on quality results, but little to no improvements are achieved.
  3. By the time data is collected, analyzed, and presented to leadership, it is outdated.
  4. Management and staff do not trust the data or place any priority on it.
  5. Results are often perceived as biased and incomplete.

So, considering the high costs of quality programs and some or all these common issues, it is not surprising that many organizations question the value of their programs and often will look there for cost reduction opportunities.

The good news is that quality can absolutely provide a significant ROI free of these concerns. But for many, these programs need to be turned on their head to escape the rut that may be plaguing some legacy approaches.

Seven key areas that will turn your quality program around:
With the evolution of digital channels and omnichannel journeys, it can seem overwhelming to create a holistic quality program that is capable of realistically (and quickly) assessing the customer’s end-to-end journey. Technology plays a role, but you do not have to wait for the latest and greatest technology to make some significant and meaningful changes to your approach. With the increased number of interactions handled in digital channels, you may think your contact center quality program is of less importance, but it is actually the opposite, given that your agents are likely handling the more difficult customer transactions. Your contact center’s quality program provides one of the biggest windows into your customers’ experience and can be a treasure trove of information to your organization. It can tell you not only how your agents are handling customer transactions, but also how your policies and processes are impacting customers. Consider the following changes in approach to get more value out of your current quality program:

  1. Focus on business-level results versus agent-level –This is not to say that agent performance is not important because it absolutely is, and you should continue to monitor for agent performance. But, the biggest bang for your buck is to ensure your quality program is structured to assess and capture all issues impacting the customer including process, policy, technology, and content. It is critical to have an accurate picture of performance at the overall business level and you can only do that if your program is set up as such. For example, it is not uncommon that over 75% of reasons issues are not resolved are outside of the agent’s control, yet these are rarely captured via quality monitoring programs. These are the details of greatest importance because they will have a much bigger impact than trying to solve all performance issues via coaching. By migrating to an approach that focuses on business-level results in addition to agent-level, your forms will need to align with that approach with the ability to capture and drill down to all areas impacting performance. Your business-level approach can also be designed to capture incredibly valuable information for other parts of the organization including marketing, product development, and engineering. Additionally, you can capture key customer satisfaction indicators like customer effort and sentiment.
  2. Design your forms based on data, not opinions – One of the biggest pitfalls in many quality programs is the form. Everything starts with the form, and forms are often developed over time based on opinion versus data. There is often reasonable rationale why a stakeholder believes an attribute is critical, but when the data is analyzed to validate the importance, organizations find the attribute is not critical. There are many ways to design a form but the most important consideration is to ensure you have solid data as to why an attribute should be included on the form – does it really have a significant impact on the customer or the business? Do you have supporting data to show that? Keep your focus on the attributes that truly have a critical impact.
  3. Develop quality metrics that provide an accurate view of performance – There are multiple ways to measure quality and they all have their pros and cons. The key is to ensure your quality metrics can be correlated to other metrics such as customer satisfaction, issue resolution etc. To do this, you may need more than one overall score. For example, to correlate quality to customer satisfaction, you need a distinct “customer quality” score to compare. How you weight your attributes is also important and ties back to the previous point. If you are focused primarily on attributes that you know have a direct and critical impact on performance, then each attribute would be considered critical. If not, ensure you have solid, data-based reasoning for your attribute weightings.
  4. Implement a statistically valid sampling approach – Many organizations over-sample at the business level. To establish your business-level sample size, first determine at what level you would like a statistically valid sample. This could be at a line of business level or for a specific contact type or service journey. Then, by using a sample size calculator, you can quickly determine an appropriate sample size based on contact volume, historical defect rate, and desired precision rate. There is usually a “law of diminishing returns” with sample sizes and you reach a point where monitoring more does not greatly improve your precision rates. For business-level sample sizes, the best practice is typically to ensure a random and unbiased approach once you determine your population. But at times you may decide, by design, to sample a more targeted population depending on your objectives. For agent-level sampling, meaning the number of transactions you want to monitor for each agent for coaching purposes, it is difficult to achieve a statistically valid sample at a monthly or even quarterly level. The key is to establish criteria and then monitor low-performing and new agents more frequently. If using agent-level quality results for compensation purposes, be cautious given the low sample sizes as you may need to consider evaluating the data on a rolling basis.
  5. Organize your quality team to support your approach – There is no right way to structure a quality organization since it often depends on size, complexity, and scope of the organization. To achieve a completely unbiased, representative, and accurate view of BUSINESS performance, it is often considered best practice for a centralized quality team to conduct business-level evaluations. For agent-level evaluations, some organizations have supervisors conduct these because they are closest to their agents and should be spending a significant amount of time with agents anyway. Other organizations will have team quality agents or coaches conducting these. The most important thing to consider is to have a defined organizational structure to accomplish your objective for both business and agent-level monitoring.
  6. Develop a robust calibration approach – Calibration among quality evaluators is one of the most important components of a quality program, yet it is often one of the most challenging. It can be time consuming, difficult to measure, and cumbersome to manage. Often, quality organizations will schedule calibration sessions which consist of a meeting in which quality evaluators listen to a call or review a transaction, score it together, and then discuss it. This is not calibration. Best practice is to assign a Gauge (or the “expert”) who scores selected transactions and then require each quality evaluator to score the transaction independently. All evaluator scores are compared to the Gauge and results are then measured at the overall, evaluator, and attribute levels. Any evaluators under a certain threshold should be calibrated until they meet minimum requirements. If any attributes are scored incorrectly by multiple evaluators, that could indicate a broad scoring issue and that would warrant a meeting. The good news is that you do not necessarily need to have a meeting every time you want to calibrate – only when there are specific scoring inconsistencies that need to be addressed. This does require the ability to measure and quantify calibration accuracy for each session, and over time at the evaluator and attribute level. Once this is in place it will save time and money, while most importantly ensuring scoring accuracy and consistency.
  7. Implement a structured reporting and performance improvement process – Last, but certainly not least, once you make these changes you will have incredibly valuable business-level data which will guide your organization to changes that will have a significant impact. What you do with this data is where the rubber meets the road, so to speak. You must have a structure and approach to analyze the data, compare it to external metrics (such as customer satisfaction), consume it, and develop action plans to improve the areas that will have the greatest impact. There is no right approach, but most importantly ensure that you have the right team, resources, cadence, and closed-loop approach to deliver value and improvements.

Food for Thought: The Future of Quality

Above are some changes you can make immediately to your current quality program, regardless of where you are with your technology capabilities. Even with the potential of more sophisticated automation, these approaches allow you to build a solid quality program foundation for any solution you implement, now or in the future. The future is exciting with technology solutions evolving daily to further enhance and optimize quality programs.

So, as you develop your long-term quality strategy, particularly with a larger work-at-home (WAH) workforce and more complex customer journeys across both digital and live channels, technology solutions that support your strategy will come into play if they haven’t already. Keep in mind that some of these solutions may not be as mature or widely used yet in the quality space, but the landscape is changing quickly and should at least be understood as you continue to transform your quality program. Most notably is the use of AI. AI adoption continues to increase in the support of both customers and agents. It is quickly becoming a technology utilized for quality management and analytics.

As customer journeys become more complex and we have a more disbursed workforce, quality monitoring also becomes more complex, with the added difficulty of pulling data from multiple sources in real-time. Rana Gujral, CEO of Behavior Signals, explains in an article published on www.tdwi.org and titled How AI is Transforming Call Centers with Actionable Insights, “AI improves this situation by performing those basic functions (monitoring, analysis, and support) in real-time and at scale. Every call is monitored and recorded. Speech patterns are analyzed to determine the mood and response of both the customer and the agent on each call. The outcome is recorded and compared against the actions taken by the agent on the call. Much of this data can be delivered in real-time to the customer service agent, enabling them to respond in the most effective way. It also ensures that customers are matched with the right agent much faster.”

Real-time, customer sentiment, and predictive analytics can also be driven by AI. As discussed above, it is becoming more challenging, with human resources alone, to capture all interactions across all channels and identify the issues the customer experienced. Gujral explains that with AI, “This data can be captured and integrated into a single profile, enabling rapid assignment of the correct customer service agent who can proactively address their problem before it escalates.” Additionally, he explains that AI can capture specific data points related to customers’ vocal characteristics, call history, nature of their issue, and agent response to flag a potential issue and intervene more quickly.
In short, AI-modeling solutions can consume more data from multiple sources across the entire customer journey. This combined with the ability to interpret the customer’s sentiment and agents’ more subjective behaviors could lead to more objective insights with less bias and human effort.

Key Takeaway:

You can turn your quality program from a process perceived as a “cost of doing business” to a value-add to the entire organization. The biggest bang for your buck is ensuring you have a business-level approach that allows you to capture all issues impacting the customer and the business, including process, product, and policy deficiencies. There are cutting-edge technology trends in quality, specifically with AI and analytics, but a solid quality framework and foundation is essential regardless of the technology deployed. Technology solutions should be implemented to support and enhance your quality program, improving efficiencies and insights, but one does not exist without the other.
Bottom line – do not be afraid to shake up your quality program. It can be one of the greatest assets to the entire organization if it is designed and executed effectively.

We would love to hear from you if you have taken steps to transform your quality program. There is no perfect solution and we know our members would like to hear success stories and best practices regarding this topic that resonates with so many.