Meta has announced that it’s expanding its AI ad control options, which will give Meta advertisers more ways to define their specific ad goals to its automated systems, helping to drive variable response to your AI-powered promotions.
Which is all part of Meta’s grand plan to eventually enable fully automated campaigns, where marketers will simply input their product URL, then let Meta’s system generate ad creative, define optimal audience targeting and launch a campaign, within their chosen budget parameters.
That’s the next stage of Meta’s grand, AI-powered ad plan, with these latest refinements contributing another piece of that bigger puzzle.
First off, Meta’s adding new Value Optimization parameters, which will enable marketers to define their preferred campaign objective, based on the value of different outcomes.
For example:
- Advertisers will be able to optimize for return on ad spend (ROAS), which will optimize for higher return from your promotions, as opposed to broader reach.
- Advertisers will also be able to share information (via the Conversions API) to outline how much profit they make from different products, enabling Meta’s system to then optimize for higher margin sales in its targeting.
- Advertisers will also be able to optimize for ROAS with non-purchase event, based on the value of specific actions taken “such as getting someone to make their first purchase, or having someone sign up for a subscription.”
In other words, Meta’s adding in variable sales targets, which aren’t defined by the unit price, in order to help guide its ad targeting systems towards the best outcomes for each brands’ promotions.
Further to this, Meta’s also expanding these qualifiers to incremental and multi-touch attribution, which will also enable advertisers to share additional information on the actions they prefer their ad viewers to take.
“We are now offering the ability for advertisers to share more granular click-level attribution information with Meta (e.g., was an individual click ultimately credited with a conversion) via Analytics integrations with partners like Adobe Advertising, Northbeam, Rockerbox and Triple Whale. This will enable advertisers access to test a new Custom Attribution feature over the next year, where advertisers can incorporate insights from external measurement sources into our optimization and drive more outcomes as measured in their analytics tools.”
Finally, Meta’s also expanding the availability of “Value Rules,” which enables advertisers to set up rules in Ads Manager that assign higher value to certain types of customers.
“An advertiser who knows that a certain age group is typically more likely to be a repeat purchaser could create a rule that would have them bid more for these customers, since they know that they likely will have higher lifetime value and ultimately help their bottom line. In this situation, advertisers are directly communicating additional business information with Meta to specify how different dimensions drive value for their business.”
So, really, these are human refinements to better target Meta’s AI-powered ad systems, and give them more context with which to align your ads with the right audience.
So while Meta does indeed want to fully automate the ad creation process, it does still need human input to ensure that it’s focused on the right aspects, with additional pointers that cannot be communicated any other way.
Maybe that will become the new ad expertise, in refining AI targeting with the exact right guidance, based on external factors, to ensure optimal response.
I mean, AI systems will likely eventually be able to extract that info from your CRM as well, and Meta’s no doubt working on systems to do just that. But at this stage, these processes will need variable levels of guidance, in order to ensure they drive the best outcomes for your business.
