Home AI Meta Bets Its Future On AI And Advantage+

Meta Bets Its Future On AI And Advantage+

SHARE:
AI brain

Despite a prevailing narrative that Meta is giving up on the metaverse, it’s adamant that isn’t the case.

“There’s been a thing out there that we’re not interested in the metaverse anymore – but we are really interested in the metaverse,” said Nicola Mendelsohn, head of Meta’s global business group, speaking during a press event in New York to update reporters on Meta’s AI-powered ad products.

“But we are also really clear that it’s going to be five or 10 years before we realize the vision of what we’re talking about,” she said.

In the meantime (and in reality, as opposed to virtual reality), Meta is spending more time, effort and money on AI technology to boost performance and efficiency in its ads system.

This includes the release of three new features within the Meta Advantage suite and an AI Sandbox for testing early versions of new tools and features, such as generative AI for ads.

Growing pains

Meta is allocating an increasing percentage of the billions of dollars it spends annually on its infrastructure toward building the company’s “capacity for artificial intelligence,” said John Hegeman, Meta’s VP of monetization.

When it comes to advertising, that means developing more complex AI modeling, building a framework for faster data processing and weaving more automation into ad products, including Advantage+, which some consider to be a bit of a black box. (That whole good-performance-lack-of-transparency thing.)

Over the past six months, the number of advertisers running Advantage+ Shopping Campaigns across Meta’s platforms every week has tripled.

Automation isn’t a panacea, however.

Just ask the advertisers who inadvertently overspent their ad budgets by tens of millions of dollars across Facebook and the Meta Audience Network late last month as a result of a glitch in the platform’s automated ad systems.

Subscribe

AdExchanger Daily

Get our editors’ roundup delivered to your inbox every weekday.

But AI is Meta’s trump card for dealing with signal loss.

Advantage+ and the new AI Sandbox aren’t a specific response to Apple’s release of AppTrackingTransparency. But both will help Meta make more accurate predictions and improve performance, Hegeman said, including in “cases where we might not have access to the same level of granular data that we may have had in the past.”

Playing in the sandbox

The purpose of the AI Sandbox, which is being tested by a small number of brands like Jones Road Beauty, is to give Meta an opportunity to quickly test and learn which new AI features are helpful for advertisers.

Meta plans to slowly open the sandbox to more advertisers in July. Based on feedback, it will then integrate some of the more useful features directly into its ad products throughout the year.

The first three features Meta is messing around with in its AI Sandbox are:

  • text variation (generating multiple versions of ad copy and messaging to find the right one for specific audiences)
  • background generation (automatically creating multiple background images based on text inputs)
  • image cropping (automatically adjusting the aspect ratio of creative assets so they look good across multiple surfaces)

Which sounds cool, but also perhaps scary to people whose jobs involve copywriting and making creative assets.

Automation, automation, automation

People shouldn’t be afraid of the rise of AI, said Cody Plofker, CMO of Jones Road Beauty.

“We as marketers should be embracing it,” he said. “We shouldn’t be thinking about AI potentially taking jobs; we should be thinking about how it can allow us to do our jobs better.”

This is a point of view that directly aligns with Meta’s official message about its AI-based products, particularly Meta Advantage, which the company is continually tweaking with feature enhancements.

The three latest additions include:

  • the ability to automatically switch manual campaigns to Advantage+ Shopping campaigns with one click
  • the ability for advertisers to upload videos as part of their product catalog, which is an additional creative surface for serving Advantage+ Shopping campaigns
  • a measurement reporting function that compares the difference in performance between manual and automated campaigns

Meta is also starting to test a new product called Advantage+ audience that will look beyond a brand’s stated targeting parameters to reach customers it might not have thought to include in their original media plan.

Although that’s it for now, Meta plans to add more AI-powered tools to its Advantage suite on the regular.

“There’s always more room for us to improve these systems and make them better,” Hegeman said.

Must Read

Pictograph of graph, mug of beer

Inside AB InBev’s Strategy For Tapping Into First-Party Data

Pour one out for third-party data. These days, AB InBev’s digital marketing strategy is built squarely on first-party data.

4A’s Measurement Committee Says New Currencies Aren’t Ready For Prime Time – Yet

The 4A’s measurement committee, a working group for marketers and media buyers to discuss their opinions and concerns about video ad measurement, has some thoughts on the status of alternative TV currencies.

How Chinese Sellers Are Quietly Reshaping US Consumer Habits

American consumers are buying more and more online products directly from Chinese manufacturers. It’s an important change, though many online shoppers are unaware.

Privacy! Commerce! Connected TV! Read all about it. Subscribe to AdExchanger Newsletters

T-Commerce Vs. Shoppable TV

Television commerce, or T-commerce, is similar to shoppable TV: both refer to buying something you see on television. But shoppable TV is far more nascent – and also has different implications on attribution.

Why White Claw’s Parent Company Is Pouring Investment Into Headless Commerce

A booze brand and a “headless commerce” platform walk into a meeting with the CFO. That might sound like the setup for a punchline, but it’s just how mar tech works these days.

As MMM Rides Again, Google Finds Its Place In The Conversation With Meridian

Tracking is a mess. Attribution is broken beyond repair. IP address identity data may go the way of the dodo. Which means marketing mix modeling is back, baby!