How to Use Google Ads’ New Generative AI Image Generator

On April 16, 2024, Google introduced a new generative AI feature for Demand Gen campaigns in Google Ads. This tool lets advertisers create unique AI-generated images from text prompts, similar to DALL-E and Midjourney but tailored for advertising.

Google says this feature aims to enhance ad relevance and engagement on its platforms, including YouTube, Discover Ads, and Gmail.

What are the benefits and limitations of Google’s new AI image tool, and how can advertisers make the most of it? Find out more below!

How Google Ads Generative AI Image Feature Works

The feature is currently available to English-speaking advertisers and will soon be introduced in other languages.

To access it, go to the Ads section of an Ad group within any existing or new Demand Gen Campaign. Click on “Generate Images” in the Media section.

Enter prompts to create various images. If desired, click the ellipses on the generated images and select “Generate more like this” to create additional variations.

Strengths: What the AI Image Tool Does Well

Creating new ad images directly within the Google Ads interface is a useful feature. Here’s what stands out:

– Speed: Google’s tool generates images quickly, often within seconds, outperforming many other AI image generators.

– Correct Sizing: Images are automatically sized to fit all Google placements, including YouTube, Discovery, and Gmail.

– Convenience: Integrated within the Google Ads interface, it eliminates the need for downloading, reformatting, and uploading external images.

– Safety: The images are designed to be generic, avoiding issues like copyright problems, offensive content, or unusual results.

– Rapid Iteration: The quick image generation allows for creating specific ad-group creatives, testing variations, and brainstorming more concepts efficiently.

Weaknesses: Where the AI Image Tool Lacks

For users accustomed to ChatGPT-4’s DALL-E or MidJourney, Google’s tool may feel lacking. It appears less advanced in image quality and prompt detail capabilities.

In my tests, the tool often did not generate images for about two-thirds of the prompts I tried. This is likely due to Google’s cautious approach following previous backlash against its Gemini image generator.

Content Restrictions

Google’s generative AI tools in Ads have specific content limitations. According to the product guide:

Faces, children, or identifiable individuals
Branded items and logos
Opinions or advice
Images that might breach Ads Policy or Generative AI use policy

The restriction on faces and identifiable individuals is quite limiting. In our testing, 60-70% of prompts that work with other AI tools were either declined or yielded no results in Google’s tool. For instance, a prompt like “a picture of trees in the style of Ansel Adams” would likely be rejected.

How to Use Google Ads’ New Generative AI Image Generator Effectively

Google’s generative AI images are generally generic and not tailored to specific businesses, which can be suitable for some use cases.

When Branding Isn’t a Priority

If maintaining a specific brand image isn’t crucial, using generic AI-generated or stock images is fine. For high-profile brands like Nike or Apple, this tool isn’t suitable. However, for local businesses like pet sitters, auto repair shops, or small real estate agents, it can be quite useful.

Rapid Iterations

The ability to quickly create and adjust images is a significant advantage. For instance, a plumbing company can generate images for various services like kitchen remodels or hot water heater replacement. Similarly, a real estate company can produce distinct images for condos, luxury properties, and starter homes. This allows for more relevant and varied visuals compared to reusing the same image.

Modify Your Prompts

The effectiveness of prompts can vary. Here are some tips:

– Use shorter, less detailed prompts as the tool is designed for quick results.

– Avoid prompts related to people, brands, or specific artists, as these are often rejected.

– Aim for images that represent services or outcomes, similar to news story visuals.

Think Native Advertising

Native ads blend in with content and don’t look like traditional ads. Using AI-generated images that fit this style can be effective. For example, instead of a standard ad like “Boost Your Business with Top-Tier SEO Services!”, try a native ad style like “Tech Startup Sees 87% Increase in Leads with AI-Driven SEO Model.” This approach integrates the image into content, making it less intrusive and potentially more engaging for readers.

Final Thoughts

Google’s new generative AI feature for image creation in Google Ads could shape the future of digital advertising. This tool lets advertisers generate images from text prompts directly within their campaigns, which may improve efficiency and speed in ad creation.

The key benefit is the rapid generation of images and their seamless integration into ads, which can be useful for quick adjustments and multiple variations in campaigns. However, the tool’s limitations in image quality and content specificity might not suit all advertisers. Those needing highly detailed or custom images might find it lacking. Additionally, restrictions on certain types of content could limit creative options.

Advertisers should evaluate this technology based on their specific needs and consider using it alongside traditional methods to get the best results.