Table of Contents
Introduction
Meta, the company behind Instagram and WhatsApp, has introduced a set of new generative artificial intelligence (AI) features for advertisers. These features include image and text generators that aim to speed up creative production and provide marketers with more control over their AI experiments. The new offerings address the demand for control and adherence to brand guidelines, ensuring that the end results meet specific internal requirements.
Greater Control and Adherence to Brand Guidelines
Meta believes that striking the right balance with generative AI is crucial for building confidence in a technology that it sees as key to future growth. To address this, the company is focusing on providing advertisers with the right levels of control. Brand guidelines are important for marketers, and Meta aims to ensure that these guidelines are not violated.
John Hegeman, Vice President and Head of Monetization at Meta, emphasized the need for control during a Q&A session with reporters. He stated that brands have specific preferences and do not want them to be violated. By offering features that allow advertisers to have control over the output of generative AI, Meta aims to increase adoption and overcome ethical and legal concerns associated with the technology.
Image Generator with Asset Submissions
Meta’s image generator, launched initially, generates image variations based on assets provided by the advertiser. Instead of relying solely on text prompts, businesses can submit existing images and ask Meta to generate new variations that align with their preferences. For example, an advertiser may submit a picture of a cup of coffee against sunny farmland and request that Meta show the cup surrounded by coffee beans and lush leaves. The software will then produce several potential image ideas that could serve as replacements.
Alvin Bowles, Vice President of Meta’s Global Business Group, explained that most marketers and advertisers prefer utilizing the assets they have provided. This approach reduces guesswork and ensures that the images come directly from the agency and client.
Meta plans to expand the image generator to function off of text prompts in the future, but there is no specific timeline for this update. Additionally, advertisers can enhance their ads with text overlays using Meta’s popular typeface options. An automated image expansion tool, already used to adjust ads to fit different Meta surfaces, will ensure that the copy aligns properly.
AI-Powered Text Generation for Ad Headlines
In addition to the image generator, Meta is expanding its AI-powered text generation to include ad headlines. This update comes as a response to feedback from advertisers who wanted more diverse suggestions that better reflect their brand values and product selling points. Meta plans to improve the output of text generation as it transitions to its more advanced Llama 3 large-language models.
Global Availability and Time-Saving Benefits
Meta’s new AI features are expected to be available globally by the end of the year. The goal of these features is to reduce grunt work and free up more time for performance-driving activities, such as creative versioning. By streamlining the creative process, Meta aims to boost ad activity and ultimately increase its revenue.
Meta’s ad revenue increased by 27% in Q1, and the average cost per ad saw a 6% increase, indicating growing demand. Alvin Bowles emphasized the importance of putting time back into the hands of those focusing on strategic relationships instead of spending excessive time on transactional activities.
Striking a Balance and Addressing Ethical Concerns
Meta recognizes the need to strike a balance between automation and human supervision in advertising. While AI has been used by most Meta advertisers for years, questions around ethics, copyright, and disclosure have arisen. Generative AI has been known to produce bizarre and off-putting images, and concerns about copyright infringement and the need for disclosure have been raised.
During the presentation, Meta executives were asked about these topics. Meta has outlined its plans to label AI-generated images and is working on ironing out the rules for advertisers. The company aims to address these concerns and provide transparency in AI-generated content.
Meta’s competitors, such as OpenAI, are also making moves in the AI space. OpenAI recently announced a new Media Manager that allows artists to opt out of having their works used as training material for generative AI.
Access and Monetization
Meta will not prioritize AI-generated ads over those created by humans or boost its own AI-generated ads over those created with rival software. The company aims to prioritize what is performing best in order to drive up the competitiveness of ad auctions. The cost of accessing Meta’s AI features was also addressed during the presentation. Meta tries to keep as many of its ad offerings free as possible, but as generative AI becomes more sophisticated and demands increased computing power, this could change in the future.
Meta’s leadership has previously indicated that charging for the use of larger AI models is one way to monetize the technology. However, for now, Meta welcomes advertisers of all sizes and aims to democratize storytelling from an advertising lens.
Conclusion
Meta’s new generative AI features for advertisers aim to provide greater control and adherence to brand guidelines. By offering image and text generators, Meta addresses the demand for control and enables marketers to ensure that the end results meet their specific requirements. These features are expected to be available globally by the end of the year and aim to streamline creative production, boost ad activity, and drive Meta’s revenue. As the use of generative AI becomes more widespread, Meta is also addressing ethical concerns and working to provide transparency in AI-generated content.
