Each survey type answers a different question. NPS asks how likely a customer is to recommend you, and is usually sent on a schedule to track the relationship over time. CSAT asks how satisfied they were, usually right after a ticket is closed or a project finishes. CES asks how easy it was to get something done, which is useful after support or onboarding. Survey features depend on your Service Hub subscription.
Responses are stored on the contact record, so scores can be used in segments, workflows and reports. That is where most of the value is: a low score can create a task for the account owner, a high score can trigger a review or referral request, and trends can be compared by segment or by support team.
The common failure is not running any surveys at all, so service quality is judged by gut feel or by the loudest complaint. The second is running them without acting on the answers, which teaches customers that feedback goes nowhere and response rates fall.
Start with one survey tied to one moment, such as CSAT after each closed ticket. Route low scores to someone who will follow up within a set time, and review comments monthly for patterns.