CES, customer effort score, asks the customer one question: how easy was it to handle your issue. Usually on a 1-7 scale. The lower the effort the customer had to put in, the higher the score.
CES tends to predict retention better than CSAT in support contexts because effort is what drives customers to switch. A customer can rate a call CSAT 5 and still leave if they had to wait twenty minutes, navigate three menus, and repeat their information four times.
For AI agents, CES is the cleaner success metric because what an AI agent does well is reduce effort, no menus, no waits, no repetition. A vendor that reports CES alongside CSAT and resolution is doing the harder, more honest measurement.
Vorel surfaces CES alongside CSAT in the per-tenant analytics, segmented by vertical and by interaction type.