Recently, Meta, together with AppsFlyer and Dentsu, analyzed more than 1.1 million advertising creative variations and released a major report summarizing the key creative elements of popular ads and the actual performance data of generative AI in advertising.
This article will summarize the key points in the report for you, and specifically break down "How to use AI to improve advertising effectiveness" to help you grasp the latest trends and comprehensively optimize your advertising strategy!
Four golden rules of advertising creativity
Meta's research shows that a successful advertising creative structure is mostly inseparable from the following four core elements:
1. Vertical video (9:16) + audio output
Give priority to full-screen short videos (compatible with Reels, Stories, etc.)
Keep audio and video in sync to enhance immersion
Meta internal data shows that this format can bring 11% higher advertising reach and fully adapt to the current platform content trends.
2. Emotional narrative: Capture the user's emotions within 3 seconds
Arouse curiosity, resonance or surprise in the first few seconds
Tell a story instead of just introducing product features
Good advertising is not about "sales promotion" but about triggering an imagination of life.
3. Show real people: let users "see themselves"
Add real-life content (employees, customers, KOC, etc.)
With expressions and sounds, it can inspire more resonance than still life ads
Content with real people appearing in Reels has 8% higher conversion rate per dollar invested than content without real people !
4. Add text overlay: clearer information and more memorable points
Use subtitles, stickers, and pop-ups to enhance key selling points
Adapt to silent viewing scenarios to improve the completeness of information transmission
5. Compared with ordinary creatives, ads that use the above four elements at the same time:
16% reduction in CPA
29% increase in conversion rate
Coverage increased by 11%
How to use AI to improve advertising effectiveness? [Practical Guide]
Meta's report mentioned that more and more advertisers are using AI creative tools, and the conversion effect of AI advertising materials is indeed better.
In addition, more than 4 million advertisers use Meta's generative AI tools every month to quickly generate advertising images, videos, and copy.
So, how can ordinary advertisers or operations teams put AI to use?
Scenario 1: Use AI to quickly generate image materials
Applicable platforms : Meta advertising tool, Canva, Runway, Midjourney, Gemini image generator, etc.
Practical uses :
Automatically generate characters and scene images for product display
Extend a material into multiple styles (technological, retro, flat illustration, etc.)
Create abstract or emotional images that are difficult to photograph
Actual performance: AI image ads CTR increased by 11% and conversion rate increased by 7.6%
Scenario 2: Using AI to assist in writing copy
Applicable platforms: ChatGPT, Copy.ai, Jasper, etc.
Practical uses:
Generate multiple versions of main titles and CTAs at once (e.g. "Get your discount now" vs. "Enjoy a limited-time discount")
Develop emotional rhetoric that suits different audiences (such as rational accusation, emotional resonance, and entertaining tone)
Suggestion: After using AI to output a variety of scripts, conduct A/B testing to find out the expression that users are most likely to respond to.
Scenario 3: AI automatically generates video variants
Applicable platforms : Meta Creative AI, Synthesia, Pika, Runway
Practical uses :
Quickly generate short video ads from product images (automatically with text and dubbing)
Production of character broadcast videos (no real person required)
Advantages: Significantly lowers the threshold for content production, suitable for small and medium-sized brands and short-term demand scenarios .
Scenario 4: AI analyzes the best creative direction
Applicable platforms: Meta Advantage+, Google Ads AI, Advertising Intelligent Analysis Plug-
in Practical uses:
Analyze audience response to different creative types (clicks, stays, interactions)
Automatically recommend creative combinations with better performance
Core Value: Make decisions more data-driven and stop making creative decisions based on "guesswork".
Friendly reminder:
AI is not a tool to "replace humans" but an accelerator to "amplify human creativity".
The soul of creativity still comes from your understanding of the product and your insight into human nature, and AI only helps you to quickly implement, test and iterate.
Multi-format and multi-style advertising creative combinations: practical strategies to improve reach and conversion rates
In the current social media environment where content is changing rapidly, it is difficult for single-format ads to capture users' attention. The Meta report points out that advertisers who use multiple creative forms and information angles are clearly ahead in terms of delivery effects.
What is a "diverse creative portfolio"?
That is, in advertising, the following different types of material combinations are used at the same time:
Diverse formats : full-screen short videos, static images, carousels, dynamic stickers
Diversified information angles:
1. Emotionally driven (e.g., "This is why I changed")
2. Functional demonstration (e.g. "How to complete XX in 3 steps")
3. User evaluation (such as "Real Feedback Clip Collection")
4. Limited time offer (e.g. "30% off your order today")
Data highlights:
Reduce CPA (customer acquisition cost) by up to 32%
Newly reached users increased by 9%
Improve the efficiency of the advertising learning period and the speed of algorithm matching
Quick start tutorial: Create a high conversion ad
You can try the following steps right away to create a more attractive and converting video ad:
Step 1: Identify your goals and audience
Is it to increase downloads? Promote purchases? Get new registrations?
Who is your core audience and what do they care about?
Step 2: Script Conception (Emotion-focused)
A hook is thrown out in the first 3 seconds: "Do you have the same trouble?"
Use a real person's first-person story to show the before and after of the transformation
Use one sentence to highlight the value of the product: "This is exactly why I chose it"
Step 3: Shooting and editing techniques
Use vertical screen shooting (9:16), preferably real people
Add voice or background music to create atmosphere
Add text overlays to strengthen the information delivery of each link
Step 4: Multi-version testing
The same story, try two different emotional routes (emotional/rational)
The same copy, with different characters (young people/working people/moms)
Prepare at least 3-5 variations for each ad and test them in rotation
Step 5: Use AI tools to speed up material production
Use Meta AI to quickly generate image backgrounds or extended scripts
Make good use of automatic editing, AI subtitle generation, etc. to improve efficiency
Analyze performance data weekly and dynamically optimize the next batch of materials
Conclusion
This creative report released by Meta not only deconstructs the formula for good advertising, but also shows how AI can help brand owners create high-performing advertisements at a lower cost, with greater diversity and higher efficiency.
Have you tried using AI to generate advertising images? Is the advertising image it generates the result you want?
