How Jira Product Discovery is Revolutionizing Product Management with AI

Hey everybody. We're at Atlassian Europe and we're having a chat about Jira product Discovery with Axel Soray, who's the product management evangelist for the product. And, well, I know a lot of people know what Jira is, but I'm not sure everybody knows what Jira product discovery is.

So walk us through what is the relationship between these things and how are people using this? Thank you for having me, first of all, and really excited to be here. So for the longest time, product managers or you know, tease that are adjacent to product management teams have not really had a space to capture all the work they have to do, plan that work, prioritize that work, and share the outcome of that prioritization with the rest of their stakeholders.

Right? So we've heard doing research, uh, about three years ago from a lot of product management teams saying, JIRA is not where we should be doing this. 'cause Jira is for committed work.

Work that we know we are going to do. Like when we've decided, okay, we're ready to go into development and we're gonna code this, we're gonna build this, this work happens in Jira, but everything that happens left of that. So the strategy, the planning, the research, the discovery, they needed a space to do that, right?

A dedicated space to do that. That's what j Product Discovery is, is a place for product management team to capture their ideas, prioritize these ideas, and then share the outcome of that prioritization with the rest of the organization. What were people using before that?

Were they just kind of scribbling notes or putting it in a spreadsheet or docs? That's a great question. So spreadsheet is, you know, the default to most things, right?

Like if you don't have a tool, a lot of it happens in a spreadsheet. And the challenge with spreadsheets is a lot of times that work is out of sync. So, uh, you can imagine a stakeholder, a leader will send you a message and say, uh, hey, can you send me like the latest version of what you're working on?

You send them this spreadsheet and the moment you send it, it's like out of sync. 'cause like the teams are moving at all times. So that's one of the issues.

The other problem is like, this information is scattered across multiple systems. So even if you're using spreadsheets, you are not connecting the dots between, you know, your CRM data that might be in Salesforce or your customer feedback that might be in like Jira service management or like wherever the data is sitting. That information was scattered across multiple places.

So a lot of these teams really needed a place like, uh, it's interesting 'cause our customers call this either like a product hub or a product operating system mm-hmm. To bring in all of this information into what they sometimes call a single source of truth. So Describe exactly what it is I experienced, so I have an idea.

Yeah. And I put it in where and how does it manifest? That's great.

Um, so in a lot of ways it looks like a spreadsheet, like spreadsheet like, or, you know, kind of a, um, how would I put this? Like a, a tabular like list experience. So, uh, people are not necessarily lost when they land in Jira product discovery for the first place.

So they will look at a list of things, a list of work items, just like you could see them in Jira. But the advantage of Jira product discovery is this whole idea of views. So views are basically different ways to slice and dice that information in a way that makes sense, uh, for the stakeholders that you're trying to have conversations with.

I'll give you an example. If you are a product manager and you're working at the team multitude in your organization, you need a level of detail, which is not the same level of detail that a VP of product or SVP of product will be looking for, right? So with the same information you can slice and dice, uh, uh, basically these ideas and the amount of information that you wanna show based on who it is you wanna have a conversation with.

For example, if I wanna have a high level strategic roadmap, I will use tools like, you know, group buys to build swim lanes or filters or field management to basically create a curated view of the same information, but for a different audience. And that's the whole power of your product discovery. Alright.

Now, you did some research in this area before you launched the product and after. So what are f what's the feedback? What are folks telling you and, you know, what's next?

So The feedback has been overwhelmingly positive. Um, JIRA product discovery, uh, is one of the fastest growing products in the history of Atlassian. So we started this journey, uh, two and a half years ago.

Since then, we've had a consistent customer satisfaction score of above 80, uh, which is absolutely amazing. Uh, and we just passed that 20,000 customer mark. So in terms of growth, that's incredible for, uh, for us and for Atlassian.

Um, one of the things we're really excited for is how do we expand your product discovery in a way that supports product management workflows, uh, in a more end-to-end fashion, right? So if you think about it today, uh, product teams will come into your product discovery to uh, plan their work and, uh, use insights, for example, to, uh, have evidence of how they should prioritize their work. We wanna go further left of that and to do that, we have recently acquired a company called Cycle.

Sure. Uh, they were, they are actually an operator in the feedback management space. And basically they, uh, use AI to, uh, allow product teams to better curate and make sense of the volume of feedback they're collecting across multiple sources.

So we are building that capability into Jira product discovery so that product managers and product teams can have a high level of fidelity in their product decision making using evidence and insights. Alright, Now Atlassian is making massive investments in AI and there's a thing called Atlassian Intelligence, which is kind of a framework. How might that manifest itself inside of this experience?

So Atlassian intelligence is basically going to manifest itself in, for example, the, um, idea editor. So if you, you're writing an idea, let's say an idea is, um, strategic, uh, piece of work you are planning for next year, let's say a new AI capability, and you're building that AI capability as a product team, as you're writing the content in the idea, let's say it's a product requirements document, Atlassian intelligence will help you draft that based on all of the context that it has in memory. And, uh, it could be like pages you already have in Confluence, it could be work you've already done in Jira.

And we bring the power of the teamwork graph, which is basically all of the knowledge that Atlassian holds for your company in your workspace into that Atlassian intelligence experience. Will it maybe even suggest new ideas? I mean, how far can it go?

That's, That's a great question. Uh, right now we really focused on helping product managers, uh, to make sense of the data they're already collecting, right? So if you think about a typical product management team, they might be receiving information from customer support teams to Jira service management.

They might be receiving, uh, emails directly from customers asking to support them on something or providing feedback about, about a product. They might be receiving internal feedback from sales teams through Salesforce. So we're gonna use, uh, the power of ro o, which is our, uh, agentic platform to bring in all of that data, make sense of it, and provide insights in Jira product discovery.

And the, the way AI will help you do this, it will surface a lot of the commonality that exists around all of these different touch points. For example, let's say you have received 5,000 different pieces of feedback, it will go and identify themes across these 5,000 pieces of feedback and say, Hey, you might wanna look at, you know, this theme and this theme and this theme. Because looking at the volume of feedback that we've received, there seems to be something of interest here.

And then based on these suggestions, a product manager can then go and think, okay, I can see that customers over the last six months have provided us a huge volume of feedback around this specific topic or this specific pain point they're addressing. Let's do a deep dive into that. And this is gonna help them prioritize the work that they do in order to help customers make progress in their work.

I'm not sure you could do this, but in large companies especially, there's often an issue where somebody has an idea, but somebody actually already did that, you know, a couple of years ago. Yes. When it's dis languishing somewhere.

Yeah. Can I find out what other people have already done in the company? Yeah.

And kind of Reuse it. This is, this is already possible today, and this is nothing specific to Jira product discovery, right? This is available through Rover, which is, uh, uh, our agent platform.

It's available across all of our Atlassian products. So typically the way you would do this, and this is something that I I would do regularly, is if I, if I'm in a like large enterprise company where there are like hundreds of teams, and as you mentioned maybe, uh, you know, Michael from this like other team like way remote from where I sit in the organization has already worked on this. And easy way to find out is I'll just go in, uh, my workspace and ask Rvo, has anybody worked on this topic before?

And let's say the topic, let's say you're in financial, financial services and the topic is like personal finance management. And I would just simply ask, has anybody worked on personal finance management before? Question mark And Rover is gonna instantly go and look at all of the information it has indexed across your company, Lage, uh, over the, the history of information that it has.

And it'll quickly come back and say, this is what I can tell you about performance, uh, personal finance management, and this is the work that, you know, I can relate to it. And it will provide work tickets in jwa. It will provide, uh, conference documents, it will provide even items that you've referenced from external systems like Google Drive for example.

So as I understand it, you did a survey and you know my, one of the high points of that survey, but you've been doing this for a while. Is there anything in that survey that surprised you or that you didn't think you'd hear? Yeah, so we recently launched our inaugural 2026 state of product report.

And there are a few things that I thought were really interesting. So the number one thing is like AI allows teams today to like claim back, let's say two hours a week. Um, but the time that the teams are claiming back, they're not finding a way to invest this in strategy work.

And this I think is something super interesting. A lot of teams are still caught in busy work, so the time that they're claiming back, they're just reinvesting in a lot more busy work, right? And this is something I think we need to, uh, really think deeply about.

'cause today AI is primarily being used for incremental productivity gains. Why? I think there's a huge opportunity for teams to actually rethink how they can transform their workflows.

So it's not so much like, help me summarize this, but it's more like, help me think about how I could completely rethink how I work based on all of these new AI capability capabilities that are available to me today. So that's one of them. The second one, which I think is something that is more, uh, on the, I'd say cultural level in a lot of organizations that I speak to is that still a vast majority of engineers are brought into the discovery and the, uh, and and planning process way too late in the cycle.

Mm-hmm. So if you think about a product manager, they have an idea, they work on a a problem definition, they work on an opportunity and they start exploring that opportunity. Engineers will come by when probably they'll, they'll get involved when the product managers already have a spec of something to build.

And that is not how we think this should be done. Engineers should be brought in really early on so that they can share their view of one, what is the feasibility of like, you know, the, the different types of solutions we can think about to address this customer problem. But more importantly, there's a lot of creative element in how they address the solution.

And they've done this before and they're some of the smartest people in the room. So the earlier you bring them in, in that cycle of like thinking about how do we address customer problems, the better you, your output and your outcomes will be. You know, I've been surprised by those conversations happening so late myself because the cost of whatever it is, is gonna be determined by the engineering team.

So if I have a spec and an idea, I don't really know how much it costs to deliver, how am I gonna price it? So that's one of the aspects. But I think even going, like taking a step back and going back to basics, like there's this whole idea of diversity of thought.

Like if you're a product manager, there's a way you think about things like your brain is hardwired to like, think about things in a specific way, bringing in different, uh, types of people. And diversity of thought early on in the journey allows you to have better coverage of how to think about how you might wanna solve for a problem, right? So you think about, um, this concept of the product trio that we really try to embody at Atlassian.

So we have a product manager, a product designer, and a tech lead. These three people are gonna come together and together address how they think they're gonna, you know, solve for a customer problem. And the reason for this is that these people are accountable for different things.

A product manager will be accountable for the business value and viability of a product. A designer will be accountable for, uh, the, the usability of the product. And typically an engineer or a tech lead will be accountable for the feasibility of the product, right?

So these three areas, or or or discipline should be represented way early into the, the problem definition phase. One of the reasons for this is that all of the strategic context that later on leads to a developer actually picking up an epic and starting to work on the work and write code. That's the most valuable thing a developer needs.

Like they shouldn't find out about this new piece of work when it lends in their Jira backlog. And this is what we're trying to solve. So You touched on this and uh, just the last question, but I'd love to get your opinion on it.

So we see AI saving people time, but it doesn't necessarily transition into making the company more money necessarily. 'cause we don't change the way we work. We just kind of, to your point, do more busy work.

Yeah. What should people we really thinking about as they look at AI and new product development? What's gonna change?

I think, um, the way I think about this, and we in a lot of ways, in a lot of ways we're still early, right? We don't really know where, like, where the puck is going. Uh, the rate at which the AI space is evolving is absolutely incredible.

However, one thing I will stress on is like the reality of the work that product management teams have to do is largely anchored in customer knowledge. And I feel like a lot of companies are in this phase of going back to basics, which I think is extremely important. So yes, AI is gonna help you incrementally save time here and there, but the biggest differentiator you have right now is what a lot of people call taste or judgment, right?

So where AI can help you save a lot of time in maybe the way you, um, come up with how you gonna, you know, think about your go-to market plans for, uh, product launch or how you might think about writing your documentation for this new feature that you're launching because the AI has, the AI already has access to all of your code base. So things like that, all of this time that you're saving should really be reinvested in strategy and customer knowledge. And I cannot stress this enough because there's a lot of noise that comes with ai, right?

Like a lot of product managers are trying to keep up with the pace at which this like, uh, space is evolving. But I really think like product teams need to think deeply about where they invest their time. And radical focus means that they're there to solve customer problems.

So AI is a tool, it is an enabler, it's definitely something product teams should be raising their level of fluency on and their level of, uh, uh, knowledge on and embedding it in their workflows. But they shouldn't lose track of the prize, which is like, how do we solve customer problems in a way that makes sense for the business? All right.

Hey folks, you heard it here. AI for product teams, it's not just about working faster, it's about working smarter. Hey Axel, thanks for being on the show.

My pleasure. Thanks for having me. All right.

And we'll be back in a minute.