nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

How do you encourage your users to provide survey responses?

Mathew Lederman

For our enterprise service desk we are always looking for feedback and surveys from users. Unfortunately, the email that goes out requires multiple clicks and most users don't take the time to leave a survey let alone any comments.

Does anybody have a way to get better feedback? Native functionality, plug-ins, etc. Just looking for ideas at this point.

Accepted answers

All comments

Christopher Yen

Hi @Mathew Lederman

We're using the native functionality CSAT score, the 1-5 star response. It asks the customer for a response on every incident and service request resolution, and since we have JSM integrated with Slack it makes it easy for them to respond in either slack or email. With it being as easy as it gets, just 1 click on the star response in a Slack notification we get about a 9% response rate. It doesn't collect comments but we usually will take a look at the work item history of low responses to see what happened and what we can improve.

Mathew Lederman

@Christopher Yen does responding in Slack directly capture the results? Or is it just a link like the email? We're using native CSAT functionality as well, but getting <1% responses rate.

If it does capture results directly, is it the Assist app? Or can you share some documentation around the configuration you have in place?

Christopher Yen

Hi @Mathew Lederman

Yes our users get prompted via the Assist app in Slack, within the thread that's synced to the ticket comments. The results are captured directly within the Slack interface. 2026-09-10 13_45_17-Desktop - File Explorer.png

Settings wise it's pretty out of the box, we just have assist enabled and chat enabled with Slack workspace connected, and CSAT enabled under Customer Satisfaction in the space settings.
2026-09-10 14_01_35-Downloads - File Explorer.png

This was all I could find on the documentation, it just briefly references chat response without much detail
https://support.atlassian.com/jira-service-management-cloud/docs/what-are-customer-satisfaction-surveys-csats/

Sami Shaik

@Mathew Lederman ,

@Christopher Yen Slack route is the best one for desks whose users live in Slack, and it is worth crediting because it removes the two things that kill response rates: the channel switch and the login. If your users are on email, here is what I would check and change inside JSM before buying anything, roughly in order of how much each has moved the number on desks I have run.

1. Confirm the email is really one click. Out of the box, the survey is embedded in the Request resolved customer notification as a rating scale; the customer clicks a star in the email and lands on a confirmation page (collect CSAT feedback). If your Request resolved template has been customised to a company-branded body, check whether the scale survived the edit; a "click here to rate us" link in place of the stars is the "multiple clicks" you are describing. Also confirm the survey is on the resolved notification and not only on Customer-visible status changed, which fires more often and trains people to ignore it.

2. Kill the login wall. The confirmation page is on the portal, so a customer who is not signed in gets a sign-in step after clicking the star. For an internal desk on SSO that is usually one silent redirect; for anyone else it is the click where most of them leave. Check what a real customer sees on a phone, because that is where the survey is opened.

3. Rewrite the question. Project settings, Satisfaction settings lets you change the question phrase. "How satisfied were you with our service?" pulls fewer responses than a question about the specific outcome ("Did this fix your problem?"). Small, but it is the only copy the customer reads.

4. Send it when the customer cares. The native survey fires at resolution. If your agents resolve in batches, the email arrives days after the fix and reads as noise. If you need a different moment, Atlassian documents building your own send with automation, generating the unique survey token on the work item and emailing the link yourself (send CSAT survey via automation). Same survey, your timing and your template.

5. Set the target honestly. For an internal enterprise desk on email, a single-digit response rate on every resolution is normal, and Christopher's nine percent with a one-click Slack prompt is a good number, not a low one. Under one percent means something in 1 or 2 is broken; five to ten percent means the mechanism works and the next gain is sampling (survey a subset, follow up the low scores in person) rather than pushing everyone harder.

If you post a screenshot of what your customers currently receive, I am happy to point at which of 1 or 2 is eating the clicks.

Mathew Lederman

@Sami Shaik thanks for the detailed response.

The email is not 1-click and I believe that's one of our biggest problems. Rating the request in the email simply opens the JSM 'Please give us a rating' page. The stars selected in the email are not captured or recorded on the JSM page. Users have to rate the service a second time and then hit submit. What should be a simple 1-click response has turned into 3 clicks minimum assuming users are logged into the VPN to get past our IP whitelist and are logged into Jira and/or SSO. If not, a simple rating could take 10+ clicks.

The rest of your points are well thought out and I'll review those with my team.

Thanks again,

Mathew

Sami Shaik

@Mathew Lederman , thank you for testing it rather than taking my word, and you are right: I described the out-of-the-box flow as one click, and on your desk it is not. Atlassian's documentation says the resolved email contains the rating scale; it does not promise that the click in the email records the score on its own, and your test shows the rating page asking again, with VPN and SSO adding steps in front of it. I should have written "one click in the email, then whatever your login path adds" and I have corrected that line here so nobody reads it as a promise.

What that changes in the advice:

The clicks you can remove are the ones before the rating page, not the page itself. If the portal is behind SSO, the sign-in redirect is the step that costs you responses; a session that is already valid (same browser the user works in, not the mail client's in-app browser) usually passes through silently. Worth testing whether the survey opened from desktop mail behaves differently from mobile mail, because on most desks that is where the real difference sits.
The email itself cannot be shortened by configuration. There is no setting that makes the star in the email the final submission. The open requests are worth your vote: surfacing the CSAT survey on the portal itself, and reminders for unanswered surveys (Atlassian's own KB references the reminder request, JSDSERVER-4159 on the server side; the Cloud equivalents are linked from the same thread family).
Native Surveys are not a shortcut. JSM's newer Surveys feature (Service Collection paid plans, rolled out this year) is link-based too, so it has the same click count as the CSAT page, with richer questions in exchange (collect feedback with surveys).
Which is why @Christopher Yen route is the honest answer for an email desk with a login wall: put the prompt where the user already is and already authenticated, whether that is Slack, Teams or the portal itself via a Marketplace app that adds the rating on the request page. If the rating can be captured without a second sign-in, the response rate moves; if it cannot, no amount of copy in the email will.

If you can share what your users see after they click a star (screenshot with the URL bar), I can tell you whether it is the SSO hop or the survey page that is eating them.

Diana Ambrožuk [Kinline]

One way to reduce the friction is to show the survey directly inside the resolved request in the customer portal. When users return to read the final reply, the form is already there, so they can respond without opening a separate email or survey page. You can still send an email reminder if they haven’t responded.

Keeping the survey short and relevant also helps. A few specific questions about the support they received can give you more useful feedback than a single generic satisfaction question. Using familiar wording and your company’s tone can make it feel like a natural part of the conversation.

It’s a paid AdaptiveSurvey Marketplace app that runs on Atlassian, but it’s relatively affordable compared with third-party survey tools. It could be worth a look and try it free for a month to see if that helps to increase the response rate.

Josh

A useful way to improve this without adding another tool is to measure the funnel separately: delivered notification → rating-page load → submitted rating. Use a small set of test customers to confirm whether the email star is only a deep link or actually carries the selected value, and test desktop mail, mobile mail, and an in-app browser. If the selected star is not retained, treat that as a product limitation rather than asking users to repeat the rating. For internal users, a portal/Slack prompt where they already have an authenticated session generally removes more friction than changing the survey wording. I would also sample low scores for a human follow-up and avoid resending to everyone; a short, timely prompt usually beats more reminders.

Welcome!

It looks like you're new here. Sign in or register to get started.

Quick Links

Find more posts tagged with

cloud

jira-service-management-cloud

jira-service-management