What Contact Centers are measuring, according to the data

Metrics give contact centers a mirror to see what’s wrong and what’s right on their home turf. Beyond just answering questions like “how are we performing on FCR and AHT”, metrics can help leaders gauge the health of the organization and where there’s room to improve. 

Naturally, this inspires a lot of discussion around which metrics are the most important, which make the biggest impact, and what needs to happen to use them effectively. 

And thanks to new data gathered by ICMI, we now have even more insight into how contact center staff are using these metrics – and what their usage says about the state of the industry in 2025. Let’s dive in. 

What the most reported metrics tell us

According to ICMI’s latest survey on the state of the industry, contact centers are most likely to measure abandonment rate (85%), average handle time (84%), quality (77%), average speed of answer (76%), and agent productivity (74%). 

The popularity of these particular metrics underscores the importance of agent efficiency in responding to and resolving customer queries. This makes sense when the goal is to cover as much ground as possible, as rapidly as possible, for as little cost as possible. The idea is that this efficiency should more or less translate into a positive customer experience, but that’s not guaranteed.

As Jarrod Davis mentions in his excellent write-up, there’s a danger when benchmarks and quotas usurp customer experience as the north star. Consider what one agent shared with Davis about their experience working in a contact center: 

“One of the places I worked valued handling time above all else. You were better off not helping people and getting them off the phone. You were better off going ‘our connection is bad’ and hanging up. The people that fixed people’s problems got the most pushback. The ‘best person’ there took the most calls, but fixed nothing. His callers had to call back, but his numbers looked good on a spreadsheet.”      

Efficiency deserves to be celebrated, but not at any expense. Great performance ought to be recognized, provided it enhances – not undermines – the customer journey.

What the least reported metrics tell us

Just as we can infer contact center habits and practices by what they prioritize measuring, we can also learn a thing or two from what they’re not tracking as often.

Consider two of the least-reported metrics in ICMI’s study: deflection rate (14%) and self-service accessibility (13%). Deflection rate measures the portion of customer queries that are resolved via self-service options (e.g., online FAQs, automated IVR) independently of agent support. Self-service accessibility refers to how easily a customer can find and resolve their queries without requiring human support.

While all our research suggests that frontline human support isn’t going away any time soon, contact centers can’t ignore the reality that customers are continually seeking out self-service flexibility. Back in 2017, Harvard Business Review found that 81% of customers attempt to take care of matters themselves before reaching out to a live representative. In a study published last year, Gartner found that 73% of customers will attempt to use self-service tools even though only 14% of customer service issues are ever resolved via self-service alone.

Considering such lackluster performance, maybe it shouldn’t be a surprise that roughly 8 in 10 ICMI respondents don’t measure deflection rate or accessibility in their contact centers. We’re experiencing what Pete Humes has called a “self-service revolution” in customer expectations, and yet a majority of contact centers – the providers of this service – don’t keep tabs on how it’s performing. Is it any wonder that, when it comes to self-service, customers are often left holding the short end of the stick? 

“There’s a shift in focus to metrics that encompass the full customer journey across channels. This includes tracking the rate at which self-service tasks are completed, how many issues are resolved without needing an agent, and customer satisfaction scores (CSAT) specifically related to self-service interactions.”

– Pete Humes

Work smarter, not harder

Attrition and burnout are morale killers, and contact center agents are more susceptible to these forces than any other cohort ICMI studied. Just 54% of agents stick around past the 2-year mark – that’s not a leak; that’s a revolving door.

Contact centers want to cultivate high-performing agents, but that’s difficult to do when you’re not setting agents up for success. Only 38% of ICMI’s survey respondents say they measure agent satisfaction and well-being, as well as workforce management metrics (like utilization and forecast accuracy) that are used to allocate agent workloads and availability. 

The system isn’t set up for agents to thrive. AI, automation, and predictive behavioral analytics can help the frontline proactively triage customer issues, but in many cases agents are still regarded like firefighters rushing into a burning building — only they lack any of the equipment and training to advance in their professions. They burnout, they leave, and the search for a replacement continues. The cycle repeats.

That’s not sustainable, and it should be a wake-up call to contact center leaders to take agent satisfaction and development more seriously.    

The bottom line

ICMI’s survey on the state of the industry reminds us that metrics are important, but also imperfect. Contact centers are right to champion efficiency, speed, and productivity, but they should also be hyper-vigilant in making sure these metrics align with the mission of customer service. 

That means empowering agents without overwhelming them; investing in self-service integration where it makes sense; and incentivizing performance that focuses on long-term quality over near-term quantity.