Ask right after it happened, while they still remember what it was like.
Customer Satisfaction Score measures one specific interaction rather than the whole relationship. It is the right tool immediately after a purchase, a support ticket or a delivery, and the wrong tool for a quarterly health check.
CSAT asks how satisfied somebody was, usually on a scale of one to five. It is transactional. It has a short memory by design, and it is at its most accurate within a day of the thing it is asking about.
The usual method counts the top two answers. Add the fours and the fives, divide by the total number of responses, and read it as a percentage. Report the whole distribution alongside it, because eighty percent made of fours is a different business from eighty percent made of fives.
Within a day of the interaction. After that people are answering from memory of a feeling rather than the thing itself.
The questions, in full
Every one of them is on this page before you sign up for anything. The line under each question says what it is for, and what it is deliberately not asking.
The chart that suits the question, and a paragraph under it.
Numbers get the widget built for them. Written answers get grouped into the things people kept saying, with a count on each and every original answer still readable underneath, word for word.
Everything a results page holdsWhat goes wrong with a csat survey
- Sending it a week later. CSAT measures a moment, and the moment has gone.
- Reporting the average of a one to five scale. An average of 3.8 hides whether you have a room of fours or a split between fives and ones.
- Changing the scale between quarters. Five points to ten points looks like an improvement and is not.
- Only surveying the people who completed a purchase. The ones who gave up partway through are the ones with something to tell you.
Should I use CSAT or NPS?
CSAT for a specific interaction that just happened. NPS for the overall relationship, a few times a year. They answer different questions and most companies run both.
How many points should the scale have?
Five is the common default and it is easy to answer on a phone. What matters far more than the number of points is that you never change it once you have started.
What response rate should I expect?
Commonly between 15 and 30 in 100 when it goes out within a day of the interaction. Timing moves this number far more than anything else you control.
Often run alongside this one
CES survey
Customer Effort Score asks how hard somebody had to work to get what they needed. It is the best single predictor of whether a customer sticks around after a support contact, and it is more actionable than satisfaction because effort has a cause you can go and remove.
NPS survey
Net Promoter Score℠ is the most asked question in business, and the most wasted. The score tells you where you stand. The sentence after it tells you what to do. FreeSurvey gives you both on the same page.
Event feedback survey
Event feedback has a short half-life. Ask in the room and you get detail. Ask on Monday and you get a mood. Everything about this survey is built around being answerable on a phone, standing up, in under a minute.