usage-based pricing

enabling atlassian’s new usage-based pricing model for AI products such as Rovo and Rovo Dev, supporting revenue growth from $28M to $152M

5 months (overall project)

Timeline

5 weeks (Rovo Dev monetisation acceleration)

1 Lead Designer

1 Product Designer (me!)

1 Senior Product Manager

1 Senior Content Designer

12 x Software Engineers

Team

01 PROBLEM & GOAL

Problem: Atlassian needs to shift its monetisation model to usage-based pricing (UBP) in order to match the variable of compute-intensive AI features from it’s products and features such as Rovo and Rovo Dev

UBP enables critical goals for Atlassian, including customer fairness, continued funding for AI investment and increasing adoption of AI. However, UBP monitoring is fragmented across various products. Usage needs to be centrally managed in Atlassian Administration to avoid bill shock and increase trust.

The goal: Admins need to be able to stitch together usage, forecasts, costs and apply AI credit limits to keep their organisation running without disruption. Rovo Dev specifcially, needed to be fully monetised and integrated with this new model in just 5 weeks, ready for October Team EU ‘25 launch.

02 BUSINESS OPPORTUNITY

By creating a UBP platform for Admins, Atlassian's consumption-based pricing (CBP) strategy is projected to grow from ~$28M in FY26 to ~$152M by FY28, as more products adopt UBP for AI and other variable-cost features.

03 the process

The design process encompassed familiar stages such as qualitative research, concept testing, ideation, customer testing and integration of feedback. However, 5 weeks before launch, we needed to begin monetisation of Rovo Dev meaning accelerating design work that wasn’t meant to ship until 4 - 5 months later.

I recruited, conducted and analysed a round of qualitative discovery research with Atlassian admins, which highlighted the need for clear visualisation of usage, setting limits and effective notifications

We partnered with PM to identify the milestone plan to begin rolling out such a drastic shift and help customers familiarise themselves with new usage-based concepts

I explored several data visualisation directions for displaying data within our component library. Which visualisation would be the most effective, line, area, bar?

With just 5 weeks, I managed the implementation and execution of designing the experience for Rovo Dev monetisation, allowing Admins to set extra usage limits

I created research stimuli and concepts to help Admins rank the importance of specific UBP capabilities

I began using AI vibe-coding tools such as v0 and Figma Make to quickly communicate concepts, ideas and prototypes to our broader working group

In May 2025, we shipped an early access program (EAP) version of the UBP platform to Admins and gathered feedback quickly to begin iterating before the next milestone

I leveraged the AI (Rovo) UX writing assistant to help produce content, while working asynchronously with our Content Designer

04 FINAL SOLUTION

Admin is able to monitor all usage across AI and usage-based products in Administration

Admin can filter by number of days and specific products/instances

Admin can easily upgrade to Rovo Dev Standard if any user is exceeding their included credits

Admin can set an extra usage limit for their users to avoid disruption to Rovo Dev usage

05 BUSINESS IMPACT

By October 2025 (Team ‘25 EU), we successfully shipped the UBP platform and enabled Rovo Dev to be monetised off this new pricing model.

This now supports UBP revenue which is projected to grow from ~$28M to $152M by FY28 as more products onboard to the platform. The next milestone of enforcement is projected to begin in 2026.

Rovo Dev officially becomes generally available and is ready to be monetised via UBP from Day 1

06 LEARNINGS

  1. Getting comfortable with shipping, learning, fast: UBP had aggressive delivery timelines, meaning I had to get used to “skipping” usability testing before implementation. This was not something I was used to compared to other projects, where I always ran user interviews and testing. However, releasing an MVP and iterating while being able to pivot quickly helped validate early and confidently while still meeting hard deadlines.

  2. Using AI as a guiding hand: This marked a new era of Product Design where we were encouraged to use new AI tools to generate ideas, build prototypes and refine designs. In general, learning tools such as V0, Replit and Figma Make helped me identify where I can condense my work flow, deliver faster and jam out ideas with PM, without straying too far from what is feasible.