Your CX team measures effort. Who’s measuring how customers feel?

Every contact center knows its containment rate. Most can report average handle time down to the second. Ask whether a customer ended the conversation thinking more or less of the brand, however, and the answers get murky.

Most contact centers are still graded like cost centers: on how quickly and cheaply they resolve issues, not on what those conversations do for the customer relationship. We found this when we surveyed 531 U.S. customer experience (CX) and contact center professionals and 238 executives for our 2026 CX Observability Study, which we ran with Metric Sherpa.

When we asked operators which metrics they track, customer effort score (CES) came out on top: 62.3% track it, more than any other operational or satisfaction metric in the study. Metrics that capture how customers felt about the experience were far less common. Only 34.1% track interaction-level customer satisfaction (CSAT), 23.0% track sentiment trajectory, and 14.9% track Net Promoter Score (NPS).

Effort tells you how hard an interaction was for the customer. It doesn’t tell you how they felt about your brand afterward. If CX teams want credit for the impact their work has on the business, they need to measure how customers feel and connect it to the numbers leadership watches.

An easy interaction may not be a good one

In fact, an effort score can make a bad interaction look like a good one.

A customer can get a quick, easy answer and still hang up with a lower opinion of your brand. Maybe the policy they asked about felt unfair and nobody explained why. Maybe the agent solved the problem but made them feel like nothing more than a ticket number. The effort score looks fine, but the customer is now less likely to buy from you again, and nothing in your reporting says so.

Sentiment trajectory and friction-point measures catch what effort scores miss. Sentiment trajectory shows how a customer’s feelings shift during a conversation and over the course of their relationship with your brand. Friction-point measures show where the same problems keep coming up, so you can fix the cause instead of resolving the same complaint over and over. Together, they show whether service is building loyalty or wearing it down.

AI agents raise the stakes. When a human agent mishandles a conversation, it usually affects one customer. When an AI agent gets something wrong, it can make the same mistake in every conversation that follows, and the damage to how customers feel about your brand adds up fast.

So why aren’t more teams measuring how customers feel?

The bandwidth problem

When we asked operators what keeps them from adopting more complete measurement, four answers came in at nearly identical rates: leadership not prioritizing it (35.6%), budget or resource constraints (34.3%), competing priorities (33.9%), and a lack of internal expertise (32.6%).

Unclear return on investment (ROI) came in a distant fifth, at 19.8%. Fewer than one in five operators doubt that better measurement pays off. What’s stopping everyone else is time, money, and people.

AI can take much of the measurement work off teams’ plates. Instead of having people review a handful of calls by hand, AI-powered observability tools analyze every interaction, whether a human agent or an AI agent handled it, and report on sentiment shifts, recurring friction, and accuracy issues in aggregate. That’s what operators are asking for: 61.4% named evaluating every conversation instead of a sample as one of their most-wanted capabilities.

How to update your CX scorecard

Effort, containment, and handle time show how efficiently your contact center runs. To show what those interactions are worth to the business, start with two changes to how you measure and report on them, and give the results at least as much weight as your efficiency metrics.

Measure how customers feel, and tie it to outcomes

Connect sentiment trajectory and friction points to numbers other teams already watch, like retention, repeat contact, and revenue.

That connection is what leadership wants: 46.4% of executives rank financial and business impact linkage among their top three CX measurement priorities.

Treat accuracy as a loyalty measure

When an AI agent gives a customer the wrong answer quickly, the interaction scores well on effort while breaking the customer’s trust. Human-agent scorecards were never built to catch that, so AI agents need measures of their own. Yet 78.5% of operators at AI-using organizations currently have no metric for hallucinations or factual accuracy on their AI agents.

Tie CX metrics to retention and revenue

Operators have spent years perfecting the metrics that keep contact centers running efficiently. That rigor makes them the right people to build what comes next.

When CX can show leadership which interactions build loyalty, which ones put revenue at risk, and why, the conversation with leadership changes. Instead of defending its budget, CX gets a say in how the company keeps and grows customers.

Read the full 2026 CX Observability Study to see where your measurement practice stands, and score your team on the five-part CX Observability Index.