Hero image — Full-width screenshot of the Firefly gallery at its best. Ideally showing the masonry grid with a rich variety of AI-generated images visible. 2400×1350px recommended.
Overview
A gallery that teaches by showing
In late 2023, generative AI tools were new to most people. Adobe Firefly had launched, but awareness of what it could actually do — and confidence in using it — was low. The biggest barrier users cited wasn't the tool itself; it was not knowing how to prompt it effectively to get the results they wanted.
The Firefly UGC Gallery was my answer to that problem. I designed the first iteration of a community-powered gallery where Firefly users could share their AI-generated images alongside the prompts and settings used to create them — giving every visitor a starting point for inspiration and a faster path to mastering the tool.
The problem
Users were learning in isolation
A Firefly visitor survey found that the inability to achieve specific results was the most commonly cited reason users stopped engaging with Text to Image and Gen Fill. Users were learning through trial and error, often without any sense of what was possible.
The hypothesis: a gallery of real user-generated content, paired with visible prompts and settings, could serve two purposes simultaneously — showcasing the power of the model and teaching people how to use it.
Research validated that users engaged with the gallery in two distinct modes: inspiration (discovering what's possible) and education (learning how to replicate or build on what they saw). Both required seeing the work and the process behind it in one place.
Research or insight visual — A slide or diagram from the UGC research report, or a visual showing user pain points. Could be a screenshot of the research deck's insight summary page. 1600×900px.
User research by Adobe Design Research, April 2024 — the gallery's two primary benefits: inspiration and education.
Design decisions
The details that made it work
As design lead I owned the end-to-end experience: the overall layout and information architecture, the flow to submit images to the gallery, a way for users to view and manage their own submissions, and the metadata hover interaction that surfaced prompts and settings inline without leaving the browsing context.
Several of the most impactful design decisions came out of hard-fought technical constraints.
Masonry grid detail — Close-up of the grid showing different aspect ratios (landscape, portrait, square) displaying cleanly without cropping. 1200×900px.
The masonry grid handling mixed aspect ratios cleanly.
Hover state — Screenshot showing the metadata overlay on hover: prompt text, style settings, and controls visible over an image. 1200×900px.
Prompt and settings revealed on hover — education without leaving the browse context.
Infinite scroll so browsing felt effortless and open-ended. A masonry grid that could accommodate images generated at different aspect ratios — landscape, portrait, square, widescreen — and display them cleanly without cropping or distortion. And randomized content on each page refresh, so returning visitors weren't always greeted by the same images. That last detail mattered: a gallery that feels static stops being worth visiting.
Submission & profile
Closing the loop for contributors
The gallery only works if creators are willing to submit their work. I designed the submission flow to be as frictionless as possible — surfacing it at the right moment in the Firefly creation experience — and gave users a dedicated view to see and manage the images they'd contributed.
Submission flow — 2–3 screens showing the flow to submit an image to the gallery. Can be a Figma mockup, prototype screenshot, or annotated wireframe. 2400×1000px or stacked vertically.
The submission flow — designed to be surfaced at the right moment in the creation experience.
My submissions view — The screen where a user can see all images they've submitted to the gallery, with options to manage them. 2400×1350px.
The submissions management view — giving contributors visibility and control over their work.
Outcome
A surface that drives real impact
The gallery shipped and was received extremely well. It has become one of the most significant acquisition and activation surfaces in the Firefly product — drawing in new users and converting a meaningful percentage directly into the creation tool.
The +37% lift in return visitor rate and +31% lift in share rate demonstrate that the gallery wasn't just bringing people in — it was giving them a reason to come back. The research conducted after launch confirmed the core hypothesis: seeing others' prompts and settings genuinely helped users improve their own results, and the gallery became both an inspiration source and a learning tool for the Firefly community.
Final / closing image — A wide, polished view of the live gallery. Could be a full-page browser screenshot or a composed mockup showing the gallery in context. 2400×1030px (21:9).