What are you measuring? The one metric that makes or breaks customer experience fulfillment.

I’ve been in the contact center world long enough to see the good, the bad, and the ugly of just about every metric you can imagine. Average handle time (AHT), first call resolution(FCR), net promoter score (NPS)  . . . you name it, I’ve tracked it, optimized it, and questioned it.

Here’s what I’ve come to believe: most metrics don’t tell the whole story.

Customers want to feel seen, heard, and valued, and they want their issues to be resolved efficiently. And you can’t accurately measure that kind of connection with traditional metrics like AHT, FCR, or even NPS.

Is the answer to stop measuring? No. Because to an extent, the business adage “what gets measured, gets managed,” still holds true. We have to make sure we’re measuring the right things. We have to make sure we’re measuring the factors that contribute to experience fulfillment.

For those unfamiliar with experience fulfillment, it’s a strategic approach to orchestrating and automating front- and back-office tasks, so that customer intelligence — such as relationship, intent, sentiment, and desired outcome — stays intact across the enterprise. That’s the “how” behind experience fulfillment. The outcome is that brands actually deliver on their promises to provide exceptional customer experiences.

But how do you know if your brand is keeping its CX promises? And more importantly, if your brand is breaking its CX promises, how do you remedy this problem? Answer: You measure and optimize a metric that really matters – customer sentiment.

Let’s take a closer look at why focusing on customer sentiment is key to achieving true customer experience fulfillment.

What happens when we don’t measure the right metrics

In a previous job, I led a continuous improvement team for a large retailer. One of the problems at this retailer was that customer hold times were extremely high. When I investigated, I learned that contact center agents were being incentivized to keep their after-call work time ridiculously low.

The goals behind the incentive seemed valid: to increase productivity and get agents to handle more calls. However, to achieve faster after-call work, agents were actually putting customers on hold during the call to finish their notes. Technically, the agents were hitting the metric. But the customer experience? Terrible.

That’s what happens when you measure (and incentivize) the wrong things. You get behavior that looks good on paper but actually costs you customer loyalty – and revenue– in the long run.

In this case, measuring (and incentivizing) speedy after-call work completely backfired. A smarter strategy would be to empower faster after-call work by giving agents tools like automatic call summarization and next best actions to eliminate manual, time consuming tasks. By automating most of the after-call work, measuring it becomes obsolete.

From lackluster surveys to real insights

Sometimes it’s not what we’re measuring, but how we’re measuring it that’s the problem. Take post-call surveys for example. They’re a common tool for gauging customer satisfaction, but asking someone to rate their experience on a scale of 1 to 5 usually yields vague, unhelpful data. To truly understand customer sentiment, we need richer, more authentic feedback like open-ended responses and behavioral signals.

One of my favorite examples of this comes from a project I worked on with a third-party claims administrator. They were using a popular cloud contact center platform and had implemented post-call surveys that included your standard “rate this call from 1 to 5” kind of thing.

I looked at it and thought, “What does a 3.5 even mean? What do you do with that?”

So, I pushed the client to try something different. We added open-ended questions like, “How did you feel about this interaction?” and “What would you change?” Suddenly, we had real data that was emotional and insightful.

Breaking through survey fatigue

While open-ended questions are much more helpful than numerical ratings, these types of surveys also have their shortcomings – namely consumers are starting to ignore them.

A 2025 Qualtrics global study found that customer feedback has fallen to an all-time low. According to the survey, after a bad experience, only 32% of respondents sent feedback directly to the company, which was down 7.7% from 2021.

So, if customers won’t answer your questions, what can you do? With the claims administration client, in addition to surveys, we also ran real-time sentiment analysis. We tracked effort. We found patterns like “this was the third time I had to call about this issue,” which told us something was broken.

Another valuable feedback we can use is called conversation intelligence.

Contact centers receive unsolicited customer feedback every day. If there’s an issue or problem in the business, it’s coming through the contact center in the form of a phone call, email, text, or chat, and those interactions contain a treasure trove of customer insights – or intelligence.

Conversation intelligence uses generative AI to analyze customer conversations across channels so brands can harmonize unstructured conversation data from disparate sources into a single data model. Conversation intelligence then collates that data in dashboards that allow users to explore and act on topics, trends, and sources of customer friction at a granular level.

So, even if customers won’t answer your survey questions, you can still create a nuanced picture of customer sentiment through conversation intelligence.

Measure what matters: Promises kept

Ultimately, the key to achieving customer experience fulfillment is to prioritize and measure what truly matters: the feelings and perceptions of your customers. By shifting the focus from traditional metrics to customer sentiment, brands can ensure they are truly meeting their customers' needs and delivering on their promises of exceptional customer experiences.

‍