Google Ads with Deepmind Veo3

A concept AI-powered Google Ads integration with Veo3 that generates video campaigns from text, images, or video, and enables fast A/B testing of ad variants.

Google Ads with Deepmind Veo3

Google Ads with Deepmind Veo3

A concept AI-powered Google Ads integration with Veo3 that generates video campaigns from text, images, or video, and enables fast A/B testing of ad variants.

A concept AI-powered Google Ads integration with Veo3 that generates video campaigns from text, images, or video, and enables fast A/B testing of ad variants.

Role: AI Product Designer & Researcher

Role: AI Product Designer & Researcher

Team Size: Solo Project

Team Size: Solo Project

Timeline: 4 months

Timeline: 4 months

Tools: Figma · Veo3 · ElevenLabs · FigJam

Tools: Figma · Veo3 · ElevenLabs · FigJam

FOCUS

FOCUS

Cut video ad creation time by 90% with AI-generated ads from text, images, and video.

Enable rapid A/B testing with editable clips and campaign variants.

Cut video ad creation time by 90% with AI-generated ads from text, images, and video.

Enable rapid A/B testing with editable clips and campaign variants.

PROBLEM & OPPORTUNITY

Google Ads is powerful for distribution, but creating video ads remains slow and resource-heavy. Current options include uploading finished videos or using fixed storyboard templates, which don’t give advertisers the flexibility to create and adapt quickly.

Veo3 can generate videos from prompts and images, but today it’s limited: clips are short, inputs are narrow, and campaign context can’t be reused. Advertisers want faster creation, more control, and simpler ways to test variations.

Through user research and platform analysis, I found a clear opportunity: embedding Veo3 into Google Ads could make video ad creation faster, more flexible, and accessible for businesses of all sizes.

How might we help advertisers easily generate and optimize video ads using AI so they can test, adapt, and scale campaigns faster?

THE DESIGN SOLUTION

Google Ads flow that lets advertisers quickly create AI video ads, run A/B tests, and measure performance.

01

AI VIDEO GENERATION OPTION

Adds a clear upfront button to generate AI videos while creating a campaign.

Puts the option directly in front of the user so they can quickly decide to try video ads without searching through menus, lowering friction and making the entry point obvious.

01

AI VIDEO GENERATION OPTION

Adds a clear upfront button to generate AI videos while creating a campaign.

Puts the option directly in front of the user so they can quickly decide to try video ads without searching through menus, lowering friction and making the entry point obvious.

03

MULTI-MEDIA UPLOAD

Lets users upload images, audio, video, and music into campaigns.

Expands creative flexibility and keeps everything structured in a familiar drive-like format, making it easy to organize assets and build richer, higher-quality ads.

03

MULTI-MEDIA UPLOAD

Lets users upload images, audio, video, and music into campaigns.

Expands creative flexibility and keeps everything structured in a familiar drive-like format, making it easy to organize assets and build richer, higher-quality ads.

04

AI PROMPT BASED CREATION

Generates ads from a simple text description entered by the user.

Converts an idea into a ready-to-use ad without manual editing, helping users who aren’t experts in video production still create polished, professional results in minutes.

04

AI PROMPT BASED CREATION

Generates ads from a simple text description entered by the user.

Converts an idea into a ready-to-use ad without manual editing, helping users who aren’t experts in video production still create polished, professional results in minutes.

05

SMART CLIP EDITING

Lets users edit specific clips by chatting with the AI.

Makes iteration easy, instead of learning complex tools, users can just describe the change they want and instantly refine the ad, speeding up the feedback and improvement process.

05

SMART CLIP EDITING

Lets users edit specific clips by chatting with the AI.

Makes iteration easy, instead of learning complex tools, users can just describe the change they want and instantly refine the ad, speeding up the feedback and improvement process.

06

BUILT IN A/B TESTING

Allows adding multiple ads in a campaign to compare performance.

Gives advertisers a simple way to test variations side by side, see which ad resonates more, and confidently invest in the creative that drives the best results.

06

BUILT IN A/B TESTING

Allows adding multiple ads in a campaign to compare performance.

Gives advertisers a simple way to test variations side by side, see which ad resonates more, and confidently invest in the creative that drives the best results.

07

PERFORMANCE INSIGHTS

Displays clear metrics and comparison graphs for ads.

Shows impressions, CTR, conversions, CPA, and ROAS in an easy-to-read view, helping advertisers quickly identify winning ads and make data-driven decisions to improve ROI.

07

PERFORMANCE INSIGHTS

Displays clear metrics and comparison graphs for ads.

Shows impressions, CTR, conversions, CPA, and ROAS in an easy-to-read view, helping advertisers quickly identify winning ads and make data-driven decisions to improve ROI.

02

CAMPAIGN DASHBOARD

Shows all live, paused, and completed campaigns in one view.

Gives a quick snapshot of status and performance, including CTR and last update, so advertisers can easily track progress and see which campaigns or ad variations need attention.

02

CAMPAIGN DASHBOARD

Shows all live, paused, and completed campaigns in one view.

Gives a quick snapshot of status and performance, including CTR and last update, so advertisers can easily track progress and see which campaigns or ad variations need attention.

CONTEXT

Google Ads is one of the most widely used platforms for businesses to promote their products and services. However, creating high-quality video ads still requires time, money, and specialized skills, resources not every business can spare.

DeepMind’s Veo3 is a generative AI model that creates videos from text prompts and images. Currently, it is only accessible through Gemini and Flow, where videos are generated based on the immediate context provided. While powerful, it isn’t designed for business workflows, where consistent brand and campaign context needs to be stored and reused.

This project explored embedding Veo3 into Google Ads to make video ad creation faster, more accessible, and more resource-efficient for businesses of all sizes.

MY REFLECTION & LEARNINGS

01 | Designing Within an Existing System

I learned how to build an entirely new flow inside an established ecosystem, working with Google’s Material 3 design system so that the new features felt native, consistent, and scalable.

02 | Translating User Insights Into Features

I practiced going beyond surface-level feedback by turning user pain points into concrete flows and feature sets, making sure every addition had a clear purpose and measurable value.

03 | Keeping Complexity Simple

I discovered how to simplify a complex process like AI ad generation into clear steps that advertisers could understand, lowering the learning curve for first-time users.

04 | Thinking at Multiple Levels

I learned to design not just at the feature level (AI generation, editing, testing) but also at the campaign level, making sure everything tied back into a seamless advertiser workflow.

05 | Designing for Iteration and Testing

I saw how important it is to build in flexibility, from A/B testing ads to editing clips mid-process, so that users could experiment and improve results without starting over.

06 | Communicating New Concepts Clearly

I practiced explaining unfamiliar AI-driven capabilities in a straightforward way so users could trust and adopt them, rather than feeling overwhelmed by “black box” technology.

FINAL SCREENS

CONTEXT

Google Ads is one of the most widely used platforms for businesses to promote their products and services. However, creating high-quality video ads still requires time, money, and specialized skills, resources not every business can spare.

DeepMind’s Veo3 is a generative AI model that creates videos from text prompts and images. Currently, it is only accessible through Gemini and Flow, where videos are generated based on the immediate context provided. While powerful, it isn’t designed for business workflows, where consistent brand and campaign context needs to be stored and reused.

This project explored embedding Veo3 into Google Ads to make video ad creation faster, more accessible, and more resource-efficient for businesses of all sizes.

PROBLEM & OPPORTUNITY

Google Ads is powerful for distribution, but creating video ads remains slow and resource-heavy. Current options include uploading finished videos or using fixed storyboard templates, which don’t give advertisers the flexibility to create and adapt quickly.

Veo3 can generate videos from prompts and images, but today it’s limited: clips are short, inputs are narrow, and campaign context can’t be reused. Advertisers want faster creation, more control, and simpler ways to test variations.

Through user research and platform analysis, I found a clear opportunity: embedding Veo3 into Google Ads could make video ad creation faster, more flexible, and accessible for businesses of all sizes.

How might we help advertisers easily generate and optimize video ads using AI so they can test, adapt, and scale campaigns faster?

FINAL DESIGN SOLUTION

Google Ads flow that lets advertisers quickly create AI video ads, run A/B tests, and measure performance.

01

AI VIDEO GENERATION OPTION

Adds a clear upfront button to generate AI videos while creating a campaign.

Puts the option directly in front of the user so they can quickly decide to try video ads without searching through menus, lowering friction and making the entry point obvious.

02

CAMPAIGN DASHBOARD

Shows all live, paused, and completed campaigns in one view.

Gives a quick snapshot of status and performance, including CTR and last update, so advertisers can easily track progress and see which campaigns or ad variations need attention.

03

MULTI-MEDIA UPLOAD

Lets users upload images, audio, video, and music into campaigns.

Expands creative flexibility and keeps everything structured in a familiar drive-like format, making it easy to organize assets and build richer, higher-quality ads.

04

AI PROMPT BASED CREATION

Generates ads from a simple text description entered by the user.

Converts an idea into a ready-to-use ad without manual editing, helping users who aren’t experts in video production still create polished, professional results in minutes.

05

SMART CLIP EDITING

Lets users edit specific clips by chatting with the AI.

Makes iteration easy, instead of learning complex tools, users can just describe the change they want and instantly refine the ad, speeding up the feedback and improvement process.

06

BUILT IN A/B TESTING

Allows adding multiple ads in a campaign to compare performance.

Gives advertisers a simple way to test variations side by side, see which ad resonates more, and confidently invest in the creative that drives the best results.

07

PERFORMANCE INSIGHTS

Displays clear metrics and comparison graphs for ads.

Shows impressions, CTR, conversions, CPA, and ROAS in an easy-to-read view, helping advertisers quickly identify winning ads and make data-driven decisions to improve ROI.

MY REFLECTION & LEARNINGS

01 | Designing Within an Existing System

I learned how to build an entirely new flow inside an established ecosystem, working with Google’s Material 3 design system so that the new features felt native, consistent, and scalable.

02 | Translating User Insights Into Features

I practiced going beyond surface-level feedback by turning user pain points into concrete flows and feature sets, making sure every addition had a clear purpose and measurable value.

03 | Keeping Complexity Simple

I discovered how to simplify a complex process like AI ad generation into clear steps that advertisers could understand, lowering the learning curve for first-time users.

04 | Thinking at Multiple Levels

I learned to design not just at the feature level (AI generation, editing, testing) but also at the campaign level, making sure everything tied back into a seamless advertiser workflow.

05 | Designing for Iteration and Testing

I saw how important it is to build in flexibility, from A/B testing ads to editing clips mid-process, so that users could experiment and improve results without starting over.

06 | Communicating New Concepts Clearly

I practiced explaining unfamiliar AI-driven capabilities in a straightforward way so users could trust and adopt them, rather than feeling overwhelmed by “black box” technology.

FINAL SCREENS

Role: AI Product Designer & Researcher

Team Size: Solo Project

Timeline: 4 months

Tools: Figma · Veo3 · ElevenLabs · FigJam

FOCUS

Cut video ad creation time by 90% with AI-generated ads from text, images, and video.

Enable rapid A/B testing with editable clips and campaign variants.

CONTEXT

Google Ads is one of the most widely used platforms for businesses to promote their products and services. However, creating high-quality video ads still requires time, money, and specialized skills, resources not every business can spare.

DeepMind’s Veo3 is a generative AI model that creates videos from text prompts and images. Currently, it is only accessible through Gemini and Flow, where videos are generated based on the immediate context provided. While powerful, it isn’t designed for business workflows, where consistent brand and campaign context needs to be stored and reused.

This project explored embedding Veo3 into Google Ads to make video ad creation faster, more accessible, and more resource-efficient for businesses of all sizes.

PROBLEM & OPPORTUNITY

Google Ads is powerful for distribution, but creating video ads remains slow and resource-heavy. Current options include uploading finished videos or using fixed storyboard templates, which don’t give advertisers the flexibility to create and adapt quickly.

Veo3 can generate videos from prompts and images, but today it’s limited: clips are short, inputs are narrow, and campaign context can’t be reused. Advertisers want faster creation, more control, and simpler ways to test variations.

Through user research and platform analysis, I found a clear opportunity: embedding Veo3 into Google Ads could make video ad creation faster, more flexible, and accessible for businesses of all sizes.

How might we help advertisers easily generate and optimize video ads using AI so they can test, adapt, and scale campaigns faster?

FINAL DESIGN SOLUTION

Google Ads flow that lets advertisers quickly create AI video ads, run A/B tests, and measure performance.

Adds a clear upfront button to generate AI videos while creating a campaign.

Puts the option directly in front of the user so they can quickly decide to try video ads without searching through menus, lowering friction and making the entry point obvious.

01

AI VIDEO GENERATION OPTION

01

AI VIDEO GENERATION OPTION

Lets users upload images, audio, video, and music into campaigns.

Expands creative flexibility and keeps everything structured in a familiar drive-like format, making it easy to organize assets and build richer, higher-quality ads.

03

MULTI-MEDIA UPLOAD

03

MULTI-MEDIA UPLOAD

04

AI PROMPT BASED CREATION

04

AI PROMPT BASED CREATION

Generates ads from a simple text description entered by the user.

Converts an idea into a ready-to-use ad without manual editing, helping users who aren’t experts in video production still create polished, professional results in minutes.

Lets users edit specific clips by chatting with the AI.

Makes iteration easy, instead of learning complex tools, users can just describe the change they want and instantly refine the ad, speeding up the feedback and improvement process.

05

SMART CLIP EDITING

05

SMART CLIP EDITING

06

BUILT IN A/B TESTING

06

BUILT IN A/B TESTING

Allows adding multiple ads in a campaign to compare performance.

Gives advertisers a simple way to test variations side by side, see which ad resonates more, and confidently invest in the creative that drives the best results.

Displays clear metrics and comparison graphs for ads.

Shows impressions, CTR, conversions, CPA, and ROAS in an easy-to-read view, helping advertisers quickly identify winning ads and make data-driven decisions to improve ROI.

07

PERFORMANCE INSIGHTS

07

PERFORMANCE INSIGHTS

02

CAMPAIGN DASHBOARD

02

CAMPAIGN DASHBOARD

Shows all live, paused, and completed campaigns in one view.

Gives a quick snapshot of status and performance, including CTR and last update, so advertisers can easily track progress and see which campaigns or ad variations need attention.

MY REFLECTION & LEARNINGS

01 | Designing Within an Existing System

I learned how to build an entirely new flow inside an established ecosystem, working with Google’s Material 3 design system so that the new features felt native, consistent, and scalable.

02 | Translating User Insights Into Features

I practiced going beyond surface-level feedback by turning user pain points into concrete flows and feature sets, making sure every addition had a clear purpose and measurable value.

03 | Keeping Complexity Simple

I discovered how to simplify a complex process like AI ad generation into clear steps that advertisers could understand, lowering the learning curve for first-time users.

04 | Thinking at Multiple Levels

I learned to design not just at the feature level (AI generation, editing, testing) but also at the campaign level, making sure everything tied back into a seamless advertiser workflow.

05 | Designing for Iteration and Testing

I saw how important it is to build in flexibility, from A/B testing ads to editing clips mid-process, so that users could experiment and improve results without starting over.

06 | Communicating New Concepts Clearly

I practiced explaining unfamiliar AI-driven capabilities in a straightforward way so users could trust and adopt them, rather than feeling overwhelmed by “black box” technology.

FINAL SCREENS