Design Contact Center Processes to Benefit from Analytics
April 01, 2014
April 01, 2014
In the last post, we discussed four components for designing processes based on insights from analytics used to improve the customer experience and deliver on contact center operational goals.
In this post, we’ll take a look at how to design processes to benefit from analytics.
One of the challenges contact center operations face with process development is that they focus solely on the outcome. Outcomes, alone, won’t really help you accomplish continuous improvement. This is because the inputs and the drivers for the outcomes are the most important elements in process design. You need to know what happened along the way to determine if processes are achieving the outcomes in the best way possible—for your customers, as well as your company.
Three Buckets for Process Design
A process workflow is based on three buckets of information that can be used in differing ways depending upon the goals for the process.
Inputs: This bucket includes insight about the call type and most of the data can be found in your CRM system or via product registration. The situational aspect will be related to the reason they called and will overlap with drivers.
Drivers: In this bucket are factors that drive the customer’s experience and satisfaction. There are a variety of data sources for determining the drivers for customer satisfaction, including call recordings, call statistics, CRM records, speech and text mines and social media.
Results: After all is said and done, the results must be measured in accordance with the outcomes the process was designed to achieve. The data sources for this bucket include customer surveys, renewals or new purchases and product returns or subscription cancellations.
Results are a combination of subjective and quantitative data including customer memories of their experience and the data reflected by their actions to either return the product or renew or buy more and even whether they decide to become an advocate and recommend a friend.
Small Data vs. Big Data
Much of the data collected lies unused because operations managers focus on completing transactions as cheaply and cost-effectively as possible, based on conventional wisdom or intuition. The truth is that you don’t need to use all of the data. Trying to do so can get very messy. Instead, start with small data that’s more easily managed.
Create a hypothesis, test your theory and identify which data helps you to connect the dots. Pick the sources you think are most likely to have influence and develop a closed-loop experience to help you figure out what’s happening across each of the three buckets that are included in the process design. It’s important to learn what the customer’s experience along the way that causes the resulting outcome. This is the only way to improve the process design.
With all the hype around big data, people believe that the only way to discover useful insights is to boil the ocean. But in our work we’ve found that analyzing 50 randomly selected calls representing the same issue, for example, can provide usable insights. Using data is really about resisting the squeaky wheel syndrome that can cause money and resources to be spent on fixing a process that will not yield business benefit. It’s more important to focus on fixing the systemic issues that will collectively have a higher impact and return on investment.
As evidence that small data works, consider one electronics company that discovered 30% of their support calls were for in-progress repairs. The insight helped them work cross-functionally to redesign workflow processes to help eliminate the frustration customers experienced in trying to get to the right department for resolution of their issues.
In the next post we’ll discuss how to use analytics to increase sales by improving the customer experience.