top of page

Setting North Star Vision for Ads Creation

Shaping a Long-Term Vision for Responsible AI Creation

Role

Product design lead

Areas

Mindset Shifting
Responsible AI at Scale

North Star Design Vision
Strategy

Systems Thinking

Jobs to Be Done

Timeline

3 months

From Builders to Editors

Effective ads shouldn’t demand expertise. By starting with pre-assembled creatives and refining them with AI-powered recommendations, advertisers can focus on decision-making—not construction.

10.png
Apple Macbook Pro 15_ (2).png

Vision delivered

The vision excited the leadership and was assigned to various teams. Big ideas, such as the universal creative builder, were included in the 2024 operational plan. Design patterns like previews, editing experiences, and side sheet frameworks were integrated into ongoing projects (e.g., AI-generated image editing).

Background

Enabling marketers to create effective ads is a major growth opportunity for Amazon Ads (visuals contribute up to 50% of overall ad performance). However, many advertisers, especially small businesses, find it hard and expensive to create good ads. For example, only 14% of Sponsored Display and 42% of Sponsored Brands advertisers use lifestyle images, even though they improve performance. Larger businesses also face challenges with following regulations and understanding how their ads are performing, which can slow down their campaigns.

Group 633560.png

Sponsored Display image ads example

Group 633561.png

Current Creative building flow

Approach

Understanding User Mindset and Pain Points​

​

This vision was guided by 5 advertiser research studies conducted over the last year, feedback from 15 customer interviews, asset adoption data, current industry trends and design workshops with different teams.

Group 1664677514.png

User research, design workshop and competitor audit, 2023

Group 633562.png

Quotes from advertiser interviews, 2023

Understanding User Workflows and Goals​​

​

To understand advertisers’ work flows and their goals. We started with a JTBD workshop with researchers.

Screenshot 2024-10-02 at 10.45.38 PM.png

Job to be done exercise, Feb 2023

Initial Explorations on Preview and Editing​

​

I started with the core preview and editing experience. Currently the preview and editing controls are pretty siloed. It presented us great difficulty in scaling up. I focused on exploring how integrated preview and editing can be without overwhelming users who usually are not design experts.

Group 63356.png

Initial explorations

Evolution – AI Drives the End-to-End Flow​

​

Guided by user research insights and rapid technological advances, I leveraged AI to significantly simplify and optimize the ads creation experience.

Group 633563.png

AI-powered ads creation flow

The Pivot

Embracing AI, I got taken away on the asset generation experience. While advertisers no longer worry about sourcing assets, creating an ad still requires multiple steps.

Revisiting advertisers’ end goals, my team and I made
a major pivot—from a builder mindset to an editor mindset. What if users could create a complete ad with one click, then edit it as needed?

Group 633565.png

1-click creative experience

Final Vision

Creative to campaign with 1 click

Refresh creative with 1 click

What’s Next

We delivered the vision in June 2023, redefining the role of ad creatives in campaign building. Instead of being an afterthought, ad creatives became the campaign's centerpiece. With AI, advertisers now effortlessly receive fully assembled creatives tailored to their goals, eliminating barriers like resource constraints or lack of design expertise.

The vision sparked excitement across Amazon Ads, influencing seven verticals and shaping Amazon’s yearly operational plan. It also guided the standardization of interaction patterns for advertiser experiences.

bottom of page