A brass pneumatic tube in an old office at dawn, a glowing capsule loaded in its open hatch and loose notes beside it. A person sorts notes at a table by the rainy windows.

Feedback analysis

Feature requests in Jira: turn reviews into one clear issue

Feature requests in Jira, from app reviews and tickets: put matching requests together, write one Jira issue in the customers' words, then send it.

Nirupa Barad

5 min read

Share

A review asks for a way to export reports. Last week a support ticket asked for the same thing in different words. Neither one reached Jira. To get feature requests in Jira, group the reviews and tickets that ask for the same thing. Then write one Jira issue (the ticket your team plans work from) for that group, with the customers' own words in it.

A request that never reaches Jira is never weighed against the rest of your work, so it is never planned.

Key takeaways

  • Write one Jira issue per request, not one per review. Group the reviews and tickets that ask for the same thing first.
  • A good feature request names the problem, not only the feature. Put the customers' own words in the description.
  • Before you send it, check that every review in the group really asks for the same thing.
  • A review is not a vote. How often something is asked tells you where to look, not what to build.

What are feature requests in Jira?

A feature request is a customer asking for something your product does not do yet, or does not do well enough.

Feature requests in Jira are those asks written up as Jira issues. Then your team can plan them next to bugs and other work.

The hard part is not Jira. It is getting from scattered reviews to one clear issue.

Why do requests in reviews never reach Jira?

Because nobody owns the step in between. The app review lands in the App Store. The ticket lands with support. The survey comment lands in a spreadsheet.

Each one says the same thing a different way. "Can I get this as a spreadsheet?" and "Please add CSV export" are one request. (A CSV is a plain spreadsheet file.) Read one at a time, they look like two small asks.

So someone has to read them all, notice they match, and write it up. On a busy week, nobody does.

How do you turn feature requests from app reviews into one Jira issue?

Do it by hand once, so you know what good looks like.

  1. Copy the last month of app reviews and support tickets into one spreadsheet. One row per review or ticket.
  2. Mark every row that asks for something new. Skip praise and plain bug reports. "The app crashes" is a bug. "I wish I could export my data" is a request.
  3. Put matching asks together. "Export to Excel", "download my history" and "CSV please" go in one group. Name the group in the customer's words, such as "Export my data."
  4. Write one Jira issue per group. Not one per review. Ten matching reviews make one issue, not ten.
  5. Paste three customer sentences into the description. Pick ones that say why they want it. "My accountant needs a CSV" says why. "Add export" does not.

Our step-by-step guide to customer review analysis covers the reading part in more depth.

How do you write a feature request in Jira?

Atlassian's own guide says good requests "clearly describe the problem, provide context, suggest solutions, and highlight benefits" (Atlassian). For a request built from reviews, that maps like this.

Part of the Jira issue What to write Example
Summary (the title) The problem, in plain words "Customers cannot get their data out of the app"
Description What they are trying to do, and why "They want to share monthly reports with an accountant"
Customer words Three short quotes, copied exactly "I retype everything into a spreadsheet every month."
Where it came from Which channels raised it App Store reviews, support tickets
Possible fix One idea, marked as a guess "A CSV download on the reports page"

Lead with the problem, not the feature. "Add CSV export" tells your team what to build. "Customers retype reports into a spreadsheet" tells them why, and they may find a better fix.

What should you check before you send it to Jira?

Two checks catch most mistakes.

  1. Does every review in the group ask for the same thing? "Export my data" and "back up my data" sound alike. One wants a file to share. The other wants nothing lost if the phone breaks. Split them into two groups.
  2. Is it a request or a bug? "Export stopped working after the app update" is a bug. It goes to Jira too, but as a bug, not as a new feature.

If a review does not fit, take it out. A group that mixes two asks becomes a Jira issue nobody can build.

Is one review a vote?

No. A request that comes up often tells you where to look. It does not tell you what to build.

A review only comes from someone who chose to write one. A few people can sound loud, and a customer who leaves without writing tells you nothing. Use how often a request appears as a reason to look closer, then decide with what you know about your product.

How Metricsense does this

Each step starts with the problem, then says what Metricsense does about it.

  1. Your requests are spread across the App Store, Google Play, tickets and surveys. Metricsense connects to the App Store and Google Play. Tickets and survey answers come in as a CSV, a plain spreadsheet file.
  2. You do not know which reviews are requests. You ask in plain words, such as "What features are customers asking for?" Metricsense calls a question like this an AI Insight. You do not sort anything into groups by hand.
  3. The same ask is worded ten different ways. Metricsense reads reviews, tickets and surveys together and groups them by reason. It calls these groups themes. One request is counted once, wherever it was raised.
  4. Your team will not act on a summary they cannot check. Every finding links to the exact customer quote behind it. Read the quotes before you send anything.
  5. The request still has to reach Jira. Send a finding from Metricsense to Jira, or to GitHub Issues if your developers plan work there.
  6. You want to know if a request keeps coming back. Trend tracking shows whether a request is growing or fading over time. Alerts by email or Slack tell you when a problem comes up again.

Here is what the question in step 2 might return. Illustrative data, not actual customer data.

Export my data (most common)

  • "I retype my monthly report into a spreadsheet. Please let me download it."
  • "Love the app, but my accountant needs a CSV."
  • "Is there any way to get my history out?"

Shared accounts for families

  • "My partner and I both track spending. We need one shared account."

What you learn from this: three customers, in three different words, have the same problem. That is one Jira issue, not three.

What stays with a person?

Deciding what to build stays with your team. Metricsense shows what customers asked for and their exact words. It does not know your plans, your costs or which customers matter most to you.

Telling customers stays with a person too. On the App Store, when you respond to a review, Apple says "the customer is notified and can update their review in response" (Apple). When the feature ships, reply to the people who asked.

For checking any AI answer before you trust it, see how to analyze customer feedback with AI.

Get your customers' requests into Jira

Connect your app store or upload a CSV. Ask what features customers want, read the quotes, then send the clearest one to Jira. Start free with Metricsense.

Share this article

Back to the blog