Call centers serve customers in many industries and service sectors, including healthcare and insurance, often in the context of claims. These may refer to property damage such as a car crash or theft, a house fire, or illness or damages suffered by the person, such as a serious injury or even death. This post focuses on Lean considerations and other process improvement (PI) aspects that may lead to better customer service.
A Lean take on call handling
It is well known that call center staff follow prepared scripts in trying to answer customer queries and respond efficiently to people who are often angry, scared, or stressed due to unforeseen events such as those mentioned above. Looking at this aspect of call handling, it is easy to spot potential deficiencies in customer service, to wit:
- overly long or unstructured scripts
- scripts that require the customer to answer questions that are not applicable to their specific situation, because the script is either too structured or too generic
- forcing the customer to stay on the line while a rep becomes available
- not having the customer information immediately at hand when the conversation starts
- making the customer repeat information they already shared with another handler on a previous call or even on the current call, as they are transferred.
Lean teaches us that there are several categories of waste potentially incurred when performing a task. By definition, waste does not add value and we should do what we can to eliminate it. Types of waste include overproduction, excess movement, defects, waiting, inventory, and non-utilized talent, among others. In our case, all of the above examples have non-value-added (NVA) steps involving, at the very least, defects, extra-processing, and/or waiting. Transferring a call without much thought to a handler whose skills do not allow them to properly service the customer while the a handler with the right skills sits idle is, additionally, an instance of talent not utilized.
Potentially wrong fixes
Much of the corrective focus by management tends to be on retraining staff to follow deficient scripts that are judged to be too time consuming to redevelop. Another tack often tried is brute-forcing a ‘solution’, by throwing more bodies at the problem to try to reduce wait times in the face of growing work volume. Doing this is likely to be insufficient and results in outcomes that are just as poor but that come at a higher cost. For example, the cost of having to train new handlers becomes an additional, distracting task for experienced but potentially overworked call center staff. The alternative, classroom learning, is unlikely to be good enough to allow a new rep to hit the ground running. There is also the additional complexity of coordination overhead due to more resources on the floor of the call center, and this is complicated by the potential unavailability of more senior staff on unpopular night shifts or weekend rotations. As work volume grows, the temptation exists for management to skimp on training and treat every staff member the same as to proficiency, even when individual capabilities clearly indicate this is not the case and the need for something else exists.
Sources of error
As we have seen, not having the right training and tools can negatively impact call center operations. A more subtle source of errors and poor service comes from processes that are not well thought out, including not assigning the work to reps with different degrees of skill according to its complexity, not capitalizing on handlers who may be better at managing their queues, and making senior staff who are specialists at handling certain very complex situations be on the phone as ‘first responders’, a task to which other specialists may well be better matched. The impact of constantly having to refocus on the work after a call has been wrapped up leads to repetitive setup time, cognitive overload, and reps feeling weary. From there to making mistakes or sub-optimal decisions is but a short step. These mistakes may lead to being out of compliance as well as failing to treat or compensate the victim of the damage accurately and in timely fashion.
Ways to try address these issues include paying great attention not only to both staff and complaint tiering, but also to scheduling and leveling the workload in the sense of calls handled per person, and to performing proper call (re)routing when the right rep is not available and response time targets are not being met, which may end with calls being assigned willy-nilly to anyone on the shift. I mentioned already the skill some reps have in managing their queues, and these ‘heuristics’ as to prioritization should be studied and used to teach others how to be more efficient, something that is separate from being proficient at claim assessment.
Proper metrics and the pace of change
Another aspect requiring close attention is designing good metrics. This is no easy task. For example, performance metrics should take into account the potential conflict between the number of calls handled (typically, higher is better) and the rest of the work to be done and having to do with the actual processing of the complaint or claim. It is unfair to penalize staff for what are conflicting goals set by management. The average handling time (AHT) for a call is one such heavily used and abused metric, since it does not reflect the quality achieved in servicing the customer. It is no good to be ‘quick’ if the customer is dissatisfied, just as dragging a call on for an unnecessary length of time is no guarantee of better service.
It is also important to use metrics to teach first, rather than to penalize. Metrics should be introduced gradually, starting with, for example, first-pass yield, meaning what percentage of the work done is done correctly the first time and does not need revisiting. Attention should also be paid to the time frame and significant effort involved in data collection, and to ‘fitness for purpose’, for example whether these metrics are best used to gauge individual behavior or should be employed to gain a collective team view for the purpose of, say, outlier detection. Having too many metrics is confusing to all, and as bad as not having any. It may, in fact, be worse, because the illusion of having things under control and an excess of ‘executive dashboards’ lulls us into complacency.
Time is a valuable commodity, hence the reason why times are measured and pored over from many angles. These times, however, are often at odds with one another, and if management is not specific as to priorities, imposes conflicting priorities, or does not have a handle on what quality is, the call center staff may be left in a situation of involuntarily gaming the system to get better ‘scores.’ Management may also attempt to prematurely optimize a process, typically when standardization and consistency in execution have not been reached, thereby making any ‘insights’ arrived at virtually meaningless. These faulty insights are often used to drive change at a pace that is too fast and does not allow for settling down and the proper evaluation of its effects, further compounding the lack of understanding as to what may be going on. The end result is usually customer service that does not reflect the right service values, as well as incurring unnecessary costs.
This has been a quick overview of a few of the call center issues that management needs to focus on if truly exceptional customer service with an eye on ROI are priorities.