Support ticket satisfaction survey
A monthly 'rate our support' average tells you little you can act on. This survey goes out when a help desk ticket closes and asks about that one conversation: was it fixed, how much work it took, and how the agent did. A low score points at a ticket you can open and read.
Best for: Support leads and help desk teams following up on individual tickets.
Use this template freeHow satisfied were you with how we handled this request?
A live preview of this template. Click through it. Nothing gets sent.
The questions
- 1How satisfied were you with how we handled this request?RatingThe ticket score. It's about this request, not how people feel about the company overall.
- 2Was your issue fully resolved?Single selectYes / Partly / No. Whether it got fixed matters more to people than how friendly the reply was.
- 3How much effort did it take you to get this sorted?Opinion scaleThe Customer Effort Score question. Low effort keeps people around more reliably than delight does.
- 4How helpful was the person who assisted you?Rating
- 5What, if anything, made this harder than it needed to be?Long text
- 6Is there anything we could have done to resolve this faster?Long text
Sample replies, counted
These replies are made up to show the method.
Asked: What, if anything, made this harder than it needed to be?
Had to explain the whole thing again when the ticket moved to a second agent.
Had to repeat myselffirst reply took 2 days
Slow first replyThe bot asked me five questions and then passed me to a person who asked the same five questions
Had to repeat myselfNothing, Priya sorted it in one reply.
Nothing, it was easyI sent screenshots on Monday and was asked for them again on Wednesday.
Had to repeat myselfWaited over a day just to hear back that someone was looking.
Slow first replyYour help article on refunds is out of date and sent me the wrong way.
Other
| Label | Replies |
|---|---|
| Had to repeat myself | 3 |
| Slow first reply | 2 |
| Nothing, it was easy | 1 |
| Other | 1 |
Three of seven had to repeat themselves after a handoff, more than any other answer. Make the bot's answers and earlier screenshots show up for the agent before you work on reply speed.
How to run it well
- Send it as soon as the ticket closes. A few days later you're measuring memory, not the conversation.
- Attach the ticket ID and agent to every response, so a low score opens a real ticket instead of moving an average.
- Watch the effort score as closely as satisfaction. A ticket that got fixed but took four replies still wears people down.
- Alert on low scores and reply within a day. A quick follow-up on a bad ticket wins back more than the original fix did.
- Read 'resolved' and 'satisfied' separately. Fixed but painful and friendly but unfixed need different changes.
When not to use this: To measure how people feel about your product overall, not one help desk ticket, use a CSAT survey instead.
FAQ
How is this different from a general CSAT survey?
Won't tying responses to a ticket make people less honest?
Where the method comes from
Related templates
- Demographic surveyShort, optional role, size, industry and region questions to split results.
- Post-purchase surveyCatch why they bought and what nearly stopped them while it's still fresh.
- Product feedback surveyWhat they tried to do, what worked, and the one change they'd make.
- Usability test surveyDid they finish, where they got stuck, and one ease rating per task.