Meta Advantage+: Why High ROAS Is A Red Flag

Meta Advantage+ is a promising new channel that has been creating a buzz on social media due to its ability to deliver high return on ad spend (ROAS) with minimum effort. However, whenever experienced marketers see high ROAS, the question of incrementality arises. Is Advantage+ driving new value or merely taking credit for conversions that would have happened anyway?

What is Meta Advantage+?

Meta Advantage+ is a portfolio of tools introduced by Meta (formerly Facebook) to help advertisers leverage the power of artificial intelligence (AI) and automation to optimize their ad campaigns. Key features of Advantage+ include:

• Advantage+ Shopping Campaigns: These campaigns use AI to eliminate manual steps in ad creation and automate up to 150 creative combinations at once, helping advertisers quickly identify high-performing ads and optimize their advertising budget.
• Advantage+ Creative: This feature automatically adjusts ad creative for each person viewing the ad, showing them the version they are most likely to respond to, resulting in better ad performance.
• Advantage+ App Campaigns: These campaigns are designed to improve the performance of app install campaigns by offering creative flexibility, split testing capabilities, and granular reporting insights.

Meta's investments in AI-powered discovery have contributed to creating a more relevant social experience and improving the experience for advertisers. Automation is a good thing – in my first job in 2009 we had rooms full of people manually calculating bids in Excel. Today that job doesn’t really exist, and yet everyone I used to work with got promoted into more strategic roles.

The Allure of High ROAS

Many advertisers have tested Advantage+ and the results so far have looked extremely positive. ROAS is a critical metric for brands, as it measures the effectiveness of advertising spend and directly impacts marketing decisions. The ability of Advantage+ to drive lower costs or a higher return on investment is great, especially when combined with the time saved through automation.

However I’ve been burned before by ROAS that looks too good to be true. My first experience was a decade ago when Google first pushed me to use more broad match keywords in my campaign–the result was $2,000 wasted on the keyword “face” (what most people search before it auto-completes to “facebook”), completely unrelated to the campaign for “face wash” I was running. Every experienced marketer has a similar story of automation gone wrong.

Another common example that everyone runs into is retargeting campaigns. If you just look at the reporting Meta gives you, these campaigns are always among your best performing. However because you’re targeting people who are already on your website and looking to buy, the campaigns claim credit for many conversions that would have happened anyway. That’s why it’s important to always be testing incrementality.

The ROAS-Incrementality Trade-off

As marketers, we must be aware that setting high ROAS targets can harm actual performance. You incentivize your team and vendors to target the “low hanging fruit”–people who were already in the market and likely to buy. This inflates your reporting figures and you don’t even know you were targeting the wrong group until you run an incrementality test later in the year.

Buyers in lower funnel media, such as retargeting campaigns, have already shown interest in your product or service. They are more likely to convert, but they may also be less incremental, as they could have reached the conversion stage without the intervention of the campaign. AI-driven campaigns like Meta Advantage+ and Google PMax (Performance Max) can sometimes prioritize these users, as they are easier to convert and therefore drive up ROAS.

AI algorithms are trained to optimize with the data they have, and can’t make these judgment calls about what’s good for your business. Your job as a marketer is to identify scenarios where the system might fail, and take proactive steps to minimize the impact on incrementality. It's important to strike a balance between machine learning and human judgment. This may involve testing different approaches, such as running incrementality experiments, employing holdout groups, or conducting geo-based tests through a vendor like Measured.

Testing Advantage+ for Incrementality

The closest corollary to Meta’s Advantage+ is Google’s PMax (Performance Max) campaigns. Measured actually tested Google's PMax campaigns by conducting over 50 incremental lift tests to analyze its performance across various advertisers. Their findings indicated that the incrementality percentage on PMax was generally higher than traditional Brand Shopping but lower than Non-Brand Shopping. That’s bad news for marketers who adopted PMax, though they did identify ways to improve the incrementality of results, like blocking brand terms.

Building on their work testing Google PMax, Measured has begun testing Meta Advantage+ for its brands, with a focus on validating incrementality and understanding its impact on marketing strategy. I’m looking forward to seeing what they come back with in terms of incrementality across a whole basket of their advertisers.

The Future of Advertising With Generative AI

Meta envisions Advantage+ as more than just a tool for automating small optimizations. The company has announced plans to develop a generative AI system capable of launching entire ad campaigns with minimal input from advertisers—requiring only a few copy notes and images.

While still in the planning stages, this technology could profoundly impact the advertising industry. Meta, a sleeping giant in AI, has one of the largest AI deployments in the form of the Facebook timeline. Their leaked large language model, LLaMA, rivals ChatGPT, and their Segment Anything Model (SAM) has gained attention in the AI art space. Although OpenAI (with Microsoft) and Google present formidable competition, don’t count out Zuck just yet.

There are risks in fully automating ad creation with AI, and marketers will likely struggle with a loss of control over aspects of their campaigns. Despite the allure of generative AI capabilities, successful implementation hinges on campaign efficiency and effectiveness. As Meta advances its generative AI technology, Advantage+ holds the potential to become an even more formidable advertising tool—so long as marketers remain mindful of incrementality.

Conclusion

Meta Advantage+ offers significant potential for brands seeking to optimize their ad campaigns and achieve high ROAS. However, brands must carefully consider their ROAS targets and the implications for incrementality. As Measured continues testing Advantage+ for incrementality, brands are encouraged to stay tuned for the upcoming results and insights.

As marketers continue to navigate the ever-changing landscape of digital advertising and automation, it's crucial to stay informed and validate the incrementality of your campaigns. Measured offers advanced analytics solutions and incrementality testing to help you make data-driven decisions and optimize your advertising strategy. To learn more about how Measured can support your marketing efforts and ensure your campaigns are truly incremental, visit our demo page and request a personalized demo today.