ADs Copilot
foundational UX exploration of AI-Native Advertising Experience to simplify complex workflows
Time
2024
Role
UX Designer
Impact
Scaled an advertising solution used by millions of SMBs, increasing store visibility and driving measurable growth in local traffic and conversions.
Context
As AI entered the advertising ecosystem, teams explored Copilot experiences independently
This led to inconsistent interaction patterns across tools and a fragmented AI experience, making it harder for advertisers to build a clear mental model. As a result, cognitive load increased, and trust in AI-driven recommendations and decision-making remained unclear.
Design Goal
Designing for focus, clarity, and trust in AI-assisted decision making.
CRAFTED VISUAL LANGUAGE & BRAND EXPRESSION
Shaped A Cohesive Visual Identity That Balances Clarity, Warmth, And Intelligence.
I collaborated with visual and product teams to align on a cohesive visual identity. We refined the color system with higher brightness and saturation for a more vibrant, lightweight feel. Inspired by the AI-driven nature, we introduced subtle diffused lighting and minimal textures to add warmth and clarity, while reducing visual noise and focusing attention on core content.
Interaction framework
Defined Two Interaction Models: Assistant Mode And Embedded Mode
Given varying user scenarios and product capabilities, a single interaction pattern is not sufficient. I introduced two distinct interaction models to better align with different use cases, ensuring flexibility while maintaining a consistent experience.
Activation flow between Assistant and Embedded modes
User\u2013copilot entry points
Defined Multiple Triggers To Seamlessly Initiate Copilot Interactions
To ensure flexibility across different scenarios, I introduced multiple entry points for Copilot activation, including fixed entry, contextual elements, system messages, and a persistent overlay. This approach allows users to engage with Copilot naturally within their workflow, reducing friction and improving accessibility.
Conversational guidance & expectation setting
Improved Usability By Designing Structured AI Conversations
Internal testing revealed that users struggled with open-ended interactions. By defining AI capability boundaries and guiding user input, we enabled more focused, high-quality conversations and improved overall usability.
MORE
PROJECTS
