As Meta moves towards more automation, the company recently gave marketers a look under the hood at its new AI ad targeting system, which it claims is driving better results.
🎯 Ad Targeting with Limited Data
To improve ad targeting with limited data, the tech giant introduced a new ad delivery process called ‘Meta Lattice', which uses multiple data points to forecast likely ad responses through AI and other predictive technology.
To expand its understanding of user interests and activities, the process looks at data across the Meta platform:
- News Feed
- Stories
- Reels
Previously, these elements were measured individually, but its advanced predictive models now take into account a broader range of data points to better anticipate user behaviors.
Quoting Social Media Today:
It’s basically an expanded database of all of Meta’s ad response activity, which, when cross-matched with all of the other information it has on each user, enables the Lattice system to better predict likely ad interest through more advanced mapping. That makes better use of all of the data that Meta can access to show people more relevant ads.
In addition, the Lattice system is also able to better contextualize longer-term ad exposure, and its relative impact on response.
According to Meta, this approach has already improved ad exposure quality by nearly 10%.
Andrew Foxwell on Meta Lattice
Every Monday, on the podcast, veteran Meta ad buyer Andrew Foxwell joins me to chat about the platform. I asked him what he thought of Lattice:
Andrew Foxwell: I think it's interesting, it's definitely continuing to remove a lot of control that we're going to have as an advertiser. It's basically Meta saying, “We know better than you what people want to see.”
And there are some areas where I think it could be helpful, where they're replacing backgrounds and they're swapping in images.
But I know, based on Meta's track record, that this isn't going to work out as well as they think it will, and that the ads aren't going to look as good as they're claiming they will. There are going to be errors and money spent on ads where they actually look like crap.
So that's my overall concern, but I think overall it will make buying more efficient and will show more relevant ads to people over time, which is a good thing.
Tod: I think that the push that we've seen these platforms do — whether it's Meta with its Advantage+, or Google with its Performance Max — all of these AI-based campaigns, it's not like Meta's going to give you a couple of options, like would you like this background or that background… It's going to handle it like it handled distribution of targeting before, which was “Just tell us generally what you want, and we'll go find the audiences.”
I think what they're doing, correct me if I'm wrong, is we're not even going to be able to see what these ad variations, the different images, the different copy necessarily are going to be. They're just going to try a bunch out.
Andrew Foxwell: Correct. That's basically what they've told us thus far. I've tried to get more information about exactly what it's going to be rolled out like, but it's definitely going to continue to remove control. So that says, you know, you're going to be moving a lot of your work on the landing pages. You're going to be moving a lot of your work onto the core creative assets themselves.
Our full conversation ran about 12 minutes; you can listen to it on the podcast. Subscribe wherever you get your podcasts, or go to https://todayindigital.com
Image: Meta
