Businesses need smarter AI systems to boost ads and address ongoing changes to data restrictions and use practices — meet Meta Lattice, AI Sandbox, and more new AI ad tools.
Meta has been using artificial intelligence and machine learning to power its apps and services, which include Meta’s ad system. With AI and ML advancements and ongoing changes to data restrictions and use practices, businesses need smarter AI systems to resolve challenges efficiently and quickly.
The AI Sandbox: A Must-Have Ad Tool
Meta Advantage’s latest addition to its ad automation suite — the AI Sandbox, acts as Meta’s testing ground for generative AI-powered ad tools and earlier ad tool versions. Meta aims to learn what advertisers need in ad tools by building helpful features, such as text variation, image outcropping, and background generation. In that way, advertisers can make ads more creative and engaging.
The AI Sandbox can help advertisers create more effective and efficient advertising campaigns. With the text variation feature, advertisers can test different messages and determine what resonates best with their target audience by automatically generating multiple versions of an ad’s text.
With background generation, advertisers can create attractive background images from text inputs. Advertisers can try different backgrounds, diversifying their creative assets. On the other hand, image outcropping allows advertisers to adjust creative assets to save time, resources, and effort in repurposing creative assets.
The AI Sandbox tools can automate many tasks involved in creating and running advertising campaigns, such as writing ad copy, creating creative assets, and targeting ads. This can save advertisers time and money, which they can reinvest in other business areas.
The AI Sandbox is still in beta, but Meta plans to add more features and functionality. Meta will gradually expand access to more advertisers and add these features to its ad products this year.
Meta Advantage’s New Features
Meta introduced new features to Meta Advantage so that advertisers can use AI to be more ad creative and measure ad strategies that work more effectively and efficiently.
Advertisers can switch manual campaigns to Meta Advantage+ shopping with just a click. This feature will be available in the Ads Manager when the advertiser clicks on ‘duplicate’ within an ad campaign.
In addition, advertisers can use video creatives in catalog ads. Aside from static product images in catalogs, advertisers can use customer demonstrations and branded videos to show them on Stories, Watch, Reels, and Feeds.
Meta’s new automatic report enables advertisers to compare manual sales campaigns and Advantage+ shopping campaigns. This new feature can help advertisers better understand automation’s positive performance, which is being rolled out using Meta Advantage+ shopping campaigns.
The Advantage+ audience is another helpful feature enabling advertisers to reach new audiences and drive conversions. Advertisers can use Advantage+ audience to add their audience inputs for Meta’s ad system to discover more prospects more likely to convert.
Meta Lattice: Meta’s Latest AI Architecture
Every year, Meta invests tens of billions of dollars in its infrastructure, particularly in AI capacity for ads. These investments ensure that Meta’s AI-powered tools can help advertisers boost their ads and achieve their full potential at scale – such as generative AI.
Meta advances its AI modeling using more complex ad models to improve performance and measurement. With more advanced modeling, Meta’s ad system can simultaneously learn and share data and transfer learning across multiple objectives. As a result, consumers will see great improvement in ad quality. These enhancements help businesses attain better results and create better consumer experiences.
Meet Meta’s newest AI model architecture — Meta Lattice. This AI model enhances Meta’s ad system, learning and predicting an ad’s performance across various datasets that small, siloed models previously supported.
Meta Lattice can improve ad performance holistically. This high-capacity architecture allows Meta’s ads system to deeply understand new data concepts and relationships, greatly benefiting advertisers through goal optimization. The early deployment results demonstrate improved ad performance on Instagram via knowledge-sharing across various surfaces (e.g., Reels, Feeds, and Stories) and advertiser objectives.
Open-Source AI Tool: DINOv2
When Meta released its Animated Drawings Demo in 2021, the company invited people to contribute to a dataset of amateur drawings. Participants uploaded images, verified or fixed a few annotation predictions, and received a short animation of their humanlike characters within their drawings. Over 3.2 million people worldwide, including social media content creators, visited the site. This huge success paved the way for more advancements in Meta’s animated drawing with the power of AI.
Mark Zuckerberg, Meta CEO, has recently announced the open-source AI tool release for creators on Instagram, including an animated drawing dataset and the DINOv2 model. The unique AI dataset animates amateur drawings, greatly benefiting animation researchers.
On the other hand, the DINOv2 implements visual tasks, like analyzing objects’ dimensions within images. Zuckerberg explained, “For example, we used it to map forest canopy height across different continents using satellite imagery. In the future, this could help improve medical imaging, food crop growth, and help make the metaverse more immersive.”
APAC Advertising impacts of Meta’s New Performance Boosters
Meta’s latest AI tools have the potential to revolutionize the advertising industry. By making it easier for advertisers to create more personalized and engaging ads, improve the performance of their campaigns, and save time and money, these tools can help businesses reach their marketing goals more effectively.
APAC advertisers are increasingly adopting AI-powered technologies to improve their marketing efforts. A recent International Data Corporation (IDC) study found that 60 percent of APAC marketers plan to increase their use of AI in marketing by 2023.
As a result, APAC advertisers will likely be early adopters of Meta’s new AI ad tools. These tools offer many benefits to help advertisers improve their campaign performance and achieve their marketing goals. As advertisers transition to new Meta AI ad tools, they’ll be able to overcome the challenges and changes in advertising. Meta drives advertiser performance through automation, predictive analytics, and animation with the help of advancing AI developments. The company continuously proactively and efficiently updates its AI architecture and models, adapting to the quickly evolving market landscape.