Meta introduces prolonged set of generative AI tools for advertisers, after first announcing a spread of AI features last October. Instead of just creating different backgrounds for a product image, advertisers can now also request full image variations that provide AI-inspired ideas for the complete photo, including riffs that update the photo's theme or the product being advertised.
In one example, Meta shows how an existing creative that shows a cup of coffee outdoors next to coffee beans could possibly be modified to present the cup from a special angle against lush greenery and occasional beans, creating images paying homage to a coffee farm .
This may not be an enormous deal if the aim of the image is just to encourage someone to go to an area coffee shop. But if the coffee cup itself was on the market, the AI variations offered by Meta could possibly be versions of the product that didn't exist in real life.
The feature could possibly be abused by advertisers attempting to trick consumers into buying products that don't actually exist.
Meta acknowledges that it is a possible use case, saying that with the upcoming text prompt feature, an advertiser could customize the output generated with different colours of their product, from different angles and in numerous scenarios. Currently, the “different colours” option could possibly be used to trick customers into pondering a product looks different than it does in point of fact.
As Meta's example shows, the coffee cup itself could possibly be transformed into different colours or shown from different angles, with each cup featuring its own distinctive swirl of foamy milk mixed with the new beverage.
However, Meta states that the system has strict protections in place to forestall inappropriate ad content or low-quality images from being generated. These include “pre-guardrails” to filter out images that don’t support the generation’s AI models, and “post-guardrails” that filter out generated text and image content that doesn’t meet their quality criteria or that they deem inappropriate. In addition, Meta stress tested the feature with internal and external experts using the Llama image and full ad image generation model to seek out unexpected uses after which remedied any vulnerabilities found.
According to Meta, the rollout of this feature has already begun, and in the approaching months, advertisers may also have the option to offer text prompts to regulate the variations of the image.
Photo credit: Meta
Additionally, Meta now allows advertisers so as to add text overlays to their AI-generated images, with a dozen of the preferred fonts available to pick from.
Another feature, Image Extension, also launched in October 2023, will now be available for Reels along with the feed on each Facebook and Instagram. This option uses AI to assist advertisers adapt their image assets to different aspect ratios reminiscent of Reels and Feed. The idea is that advertisers could spend less time repurposing their creative assets across different surfaces. According to Meta, the text overlay also works along with the image expansion.
One advertiser, smartphone case maker Casetify, said that using Meta's GenAI background generation feature resulted in a 13% increase in return on its ad spend. The company tested the choice with its Advantage+ shopping campaigns, where the AI features first became available in the autumn. As before, the updated AI features may also be available via Ads Manager via the Advantage+ creative.
Photo credit: Meta
Beyond images, Meta's AI could be used to generate alternative versions of the ad headline along with the major ad text, which was already supported by leveraging the unique copy. Meta says it tests this copy's ability to sound just like the brand's voice and tone, using previous campaigns as reference material. Text generation capabilities are being moved to Mets' next generation LLM (Large Language Model), Meta Llama 3.
All generative AI features can be available to advertisers worldwide by the top of the 12 months.
Outside of the AI updates, Meta also announced it will be expanding its Meta Verified subscription service for enterprises to latest markets, including Argentina, Mexico, Chile, Peru, France and Italy. The service began testing in Australia, New Zealand and Canada last 12 months.
Now Meta Verified offers 4 different tiers of its subscription plan, all with the fundamental features of a verified badge, account support, and identity theft monitoring. Higher tiers include latest tools like profile enhancements, connection tools, and more ways to access customer support.
Meta Verified will soon be expanded to WhatsApp, the corporate also said.
