The short answer
Jira Product Discovery is Atlassian's tool for product discovery. It collects insights from customers, turns them into ideas, scores them with formula fields you define and shows them as roadmaps in list, board, matrix or timeline views. The flow works when each approved idea links to its delivery work in Jira, and what ships is checked against the problem it was meant to solve, feeding the next round of discovery.
Where does the flow usually break?
You probably have the pieces already: feedback in the support tool and in sales notes, ideas in a backlog or a spreadsheet, a roadmap in slides, and delivery in Jira. Often Jira Product Discovery is already bought, and one product manager uses it.
What’s missing is the thread between them. Ideas reach the roadmap without their evidence, delivery work loses sight of the problem it was solving, and nobody checks whether the release solved it. Then sales promises a date that engineering never planned.
How do you collect insights in one place?
Bring customer feedback into Jira Product Discovery as insights linked to ideas. It connects to the tools the feedback already lives in:
- Support: Zendesk tickets and Intercom conversations.
- Sales: Salesforce.
- Chat: Slack and Microsoft Teams messages.
- Usage and in-app feedback: Pendo and Usersnap.
- Anything on the web: the Chrome extension, which turns highlighted text into an insight.
The point isn’t volume. It’s that every idea shows how many insights back it, from which sources, and what customers actually said.
Frame each idea as the job the customer is trying to get done, not the feature they asked for. “Finance needs to close the month without exporting data by hand” leads to better options than “add a CSV export”.
How do you score ideas without endless debate?
Agree a small set of fields with the team, then add a formula field that combines them. Two common choices:
| Method | Formula | Good for |
|---|---|---|
| RICE (from Intercom) | Reach × Impact × Confidence ÷ Effort | Comparing ideas across product areas |
| ICE | Impact × Confidence × Ease | Quick ranking within one team |
Jira Product Discovery’s built-in weighted score ranks ideas relative to each other, not on an absolute scale. A matrix view, with impact against effort, shows the trade-offs at a glance. The score doesn’t make the decision; it makes the reasoning visible, so the debate is about the inputs, not the loudest voice.
How do ideas become delivery work?
When an idea is approved, open its Delivery tab and create the epic or work items from there, or link work that already exists. An idea can link to any work type, and to several epics across spaces on the same site. A progress bar shows how far the linked work has got, by count or by story points.
One limit to plan around: progress rolls up one level, from the linked epic, not from its children’s children. For deeper roll-ups, use Plans in Jira Premium or an automation rule.
Slice the idea into releasable steps first. A story map, Jeff Patton’s technique, lays the user’s journey across the top and the steps to deliver underneath; the thinnest slice that works end to end, the walking skeleton, is usually the first release worth shipping.
| In Jira Product Discovery | In Jira |
|---|---|
| Insights and evidence | — |
| Idea, score, roadmap column | Linked epic and work items |
| Delivery status and progress | Sprint or Kanban board, release |
| Outcome checked | Release notes, follow-up ideas |
How do you keep delivery flowing?
Whether you run Scrum or Kanban, the same basics keep work moving: stories with clear acceptance criteria before they start, a Definition of Done the team shares, and limits on work in progress. The Kanban Guide names four flow measures: work in progress, throughput, work item age and cycle time. Work item age is the one that shows stuck work before it’s late, so look at it every week.
Dependencies between teams belong in Jira as links, visible before sprint planning, not discovered halfway through. With five teams instead of two, a cross-team view in Plans (Jira Premium or Enterprise) replaces the dependency spreadsheet and the meeting that maintains it.
How do you close the loop after a release?
Write release notes your support team and customers can read, from the work that shipped. Then check, a few weeks later, whether the release solved the problem: did usage change, did the related support conversations drop? Record what you learned on the idea in Jira Product Discovery. That’s the start of the next round.
What about AI features?
Building AI features follows the same flow, with one extra step before release: a defined evaluation, a test set and a quality bar the feature must meet, recorded with the prompt, model and data versions it was tested on. This is our practice, not a standard. Customer feedback on the feature links back to that version, so you can tell which change made it better or worse.
Jira Product Discovery has its own AI too: on the Premium plan, Rovo can summarize insights and comments, draft idea descriptions and answer questions about the view you have open.
Where do teams usually start?
Start with one product area: bring its feedback into insights, score its ideas, and run one idea through to release and back. That’s how the Product Ops Flywheel starts, with one flow traced live before rolling it out.
The SaaS Discovery-to-Delivery Check: 18 questions, about six minutes, your score right away.
Questions people ask
What is Jira Product Discovery?
Jira Product Discovery is an Atlassian product for product managers and teams to capture insights, organize ideas, prioritize them with custom fields and formulas, and share roadmaps. It connects to Jira, so ideas link to the epics and work items that deliver them, with delivery progress shown on the idea.
Who needs a paid Jira Product Discovery license?
Only creators, the people who create and manage ideas, fields and views. Contributors, who can view, comment, vote and add insights, are free, and people with Jira access are contributors by default. That makes sharing the roadmap inside the company cheap.
How is the priority score calculated in Jira Product Discovery?
With a formula field you add: a weighted score, a roll-up or a custom formula. The built-in weighted score ranks ideas relative to each other, not on an absolute scale. RICE (reach × impact × confidence ÷ effort, from Intercom) can be built as a custom formula.
Do we still need a roadmap in Jira if we use Jira Product Discovery?
Keep the outcome roadmap (now, next, later) in Jira Product Discovery, where ideas and evidence live. Keep delivery timelines and dependencies in Jira: the space timeline, or Plans if you're on Jira Premium or Enterprise. Link the two so each shows its side of the same work.
How do we stop the loudest customer from setting the roadmap?
Attach the evidence to each idea as insights, from support, sales, interviews and usage, and score ideas with criteria the whole team agreed on. A loud request then becomes one insight among many, and the score shows it.