Top 4 Ways to Reduce Agent Churn

Somewhere in the neighbourhood of 3 million customer service agents work in America, representing 2% of the US workforce. These agents are on the front lines of preserving customer relationships—often acting as the key competitive differentiator between competitors. Yet for as important as this job is in keeping customers satisfied, it faces one of the worst retention rates. For an industry that spends approximately 80% of total costs on this important workforce, roughly 1.2 million people leave the job each year.  

Pragmatically speaking, reducing the costs of agent turnover can go a long way in reducing the total cost to serve customers. So, what can today’s CX leaders do about the big retention problem? Drawing from new industry data found new industry reports, here are four clear ways we can reduce the churn.  

1. Prioritise Training 

As a 25-year veteran of the CX industry, I get it: there are immediate needs in getting agents on the floor to address gaps in service. Shortening the time of agent training seems like a must to confront the big problem that needs a fix without unnecessary delay. But unless we provide a solid foundation, agent retention and maintaining quality service will be a chronic problem. 

We need a hands-on approach. We’ve often used long manuals (that quickly become obsolete with new vendors or technology systems) to manage a substantial portion of agent training. Yet new research reveals that 77% of agents say hands-on training and shadowing is more effective than reading a manual. This finding corroborates a trend from newer generations of agents who prefer smaller doses of “classroom” training where agents progressively learn-then-do increasingly challenging aspects of the job.  

When agents receive more training, they report a more positive experience all around–92% saying that they are satisfied with their ability to help customers when they receive extensive training. Better trained agents are more engaged–they become strong brand ambassadors that solve customer problems and upsell products. 74% agreed that they feel like brand ambassadors when receiving adequate training. With new AI and automation capabilities filling out the contact centre’s solutions portfolio, we may see AI technologies geared specifically towards training the agents that reduce overall training time and cognitive load.  

As a part of agent training and management, coaching needs to be reflective of the actual work tasked in their role. However, 27% of agents said the feedback they received from managers didn’t reflect their overall work. Part of what may be driving this mismatch in the coaching and training of agents may have to do with how we’ve traditionally viewed agent performance metrics.  

2. Reconsider how agents are managed to traditional metrics

Metrics will always serve as critical insight into the quality of the customer experience, but given the growing adoption of AI and automation technologies, the contact center’s quest for continuous improvement will need to be looked at with different paradigms and lenses. As Max Ball, Principal Analyst at Forrester, explains in a recent blog post: “Contact centers need to look beyond narrow KPIs related to cost containment and take a broad look at the economics of running the contact center.” 

This is most apparent in average handle time (AHT) as a measure of agent performance. A conventional view might look at this metric as one that always needs to decrease. However, new technologies show that we should in fact look at this metric differently. Three reasons why:  

1. Conversations can be omni-channel, concurrent, and asynchronous. Especially with the greater percentage of digital messaging, where agents are handling multiple conversations simultaneously, or where customers may come and go.  
2. Complex issues take more time. IVRs and chatbots are serving their purpose in helping with self-service and low-complexity issue resolution, but this has meant that a greater percentage of agents’ time will need to be spent on the highly complex issues.  
3. You can’t rush great experiences. If it is the goal to provide a great customer experience, it would be wrong to penalise agents who spend extra time ensuring that customer needs are met.  

Ball expands in his blog post that the “first-contact resolution is an incredibly valuable metric, especially when customers’ callback satisfaction plummets and costs skyrocket. This is harder to measure than something as simple as AHT, but it’s where rubber really meets the road.” AHT could increase given technology advancements, and we should also look to different indicators for agent performance.  

The prevalence of mismatched metrics is perhaps most pronounced in how agents report abusing it: 78% of agents said they engage in some form of metric manipulation. Where customer surveys have a known problem in bias (oversampling for responses that have either an incredibly polarised positive or negative experience), and agents guiding positive/receptive customers to fill these out, new technology in speech sentiment could be another way to look at agent satisfaction. Instead of oversampling in just a few surveys, 100% of customer calls could be audited using speech sentiment analysis that helps present a fuller picture of what customers are experiencing.  

3. Simplify systems and moderate complexity for agents

As agents continue to operate in a difficult legacy environment with a “frankenstack” of poorly designed systems that are difficult to use, the more overwhelming and complex they become. 44% of agents state that the rate of technological advancement is behind the times.