Discover Panel and quick actions in Photoshop

Project Outcome

A unified and cohesive search & discovery capability, that's easily accessible, providing a seamless way for users to find and consume contextually relevant content right inside the creative application.

The Discover Panel leverages Adobe Sensei's AI and Machine Learning technology to drive suggestions, and, packages hidden, high-value Photoshop features as “Quick actions” by simplifying complex tasks into one-click actions tied to relevant tools and in-app learning.

Oftentimes, creative enthusiasts want to ‘Photoshop’ something simple, for example removing a busy background from a portrait image. Some problems that they run into:

• Not knowing where or how to start in Ps

• Being unable to find the relevant tool or feature

• Being unaware of less complex apps within Adobe’s Creative Cloud suite for their task

• Switching between a web article or YouTube video and the Ps app for guidance

• Giving up when they get stuck, or if the task is taking too long

The Discover Panel’s search capability, help content library, and curated quick actions aimed to alleviate some of these problems.

Making the panel persistent and dockable — rather than a search overlay — was a deliberate architectural decision. It positioned discovery as part of the workflow rather than an interruption to it, and it gave us a scalable surface to add capability over time without redesigning the core Photoshop UI.

A user can now navigate to, or search for the ‘Remove background’ quick action and have a background-free portrait in seconds. I lead the design and strategy work for 20 similar quick actions in 2021 which aimed to:

• make existing and new features more discoverable

• expedite creative workflows through automation

• create a learning opportunity by connecting to relevant help content, as well as feature controls for refinement

• improve app engagement through fun and high-value tasks

• make non-destructive layer edits

One-click actions that delivered an immediate, non-destructive result lowered the activation barrier dramatically and tied feature discovery directly to task completion.

A new content discovery experience requires feedback from users to

• improve the quality and relevancy of content

• reveal unmet user needs in the experience

• pinpoint opportunities to improve the experience

I was responsible for designing the feedback pattern to scale across different content types within the Discover Panel. This humble component lead me to constantly observe feedback gathering experiences around me - from dismissing a newsfeed article, to liking in-app informational content, and even longer form feedback on search results.

The final designs focused on:

• quick and simple binary responses

• optional open text fields for users to add details

• multiple choice responses when relevant

• an in-panel survey for more descriptive feedback

As an Adobe Spectrum contribution, I plan to meet all the criteria to make this an internally approved, scalable and holistic design pattern.

Learnings

This project required me to operate well beyond a traditional design scope — and that turned out to be one of its most formative aspects. Without a dedicated PM for significant stretches of the project, I took ownership of the release strategy: mapping dependencies, running feasibility conversations with engineering, managing content requirements across international teams, and establishing the team's first JIRA board to bring structure to a process that didn't have one.

What I learned is that design leadership at this level isn't just about the quality of your outputs — it's about keeping a complex, cross-functional system moving when the people and processes around you aren't stable. I got comfortable making consequential decisions with incomplete information, documenting my rationale clearly enough that stakeholders could stay aligned without needing to be in every conversation.

That experience fundamentally changed how I approach ambiguous projects. I now front-load alignment work — workshops, written decision logs, early stakeholder reviews — because I've seen firsthand how much ground you lose when you try to catch up later.

Despite a global pandemic, working from home with it’s own challenges, and collaborating across teams, and countries, we successfully launched the Discover Panel in 2021 and a total of 21 quick actions by 2022, which I’m very proud and honored to have been a significant part of.

Next steps

Three things stand out in hindsight.

First, I'd push harder for a tighter feedback loop between design decisions and usage data. We were often making roadmap calls — about which quick actions to prioritize, how to sequence releases — based on qualitative signal and proxy metrics. With more instrumentation in place earlier, those calls would have been faster and more defensible.

Second, I'd invest earlier in awareness. The Discover Panel solved a real problem, but reaching the users who needed it most — newer Photoshop users who didn't know where to look — remained a challenge throughout. The panel could only help users who found it. A more deliberate in-app onboarding or surfacing strategy would have amplified the impact of everything we built.

Third, I'd treat content and design as a single workstream from day one. In practice, content requirements often arrived late, which created rework cycles that compressed the time available for design exploration and user testing. Earlier integration would have improved both the quality of the quick actions and the efficiency of the process.

View more work: