Social Rocket
Meta is pushing its advertising platform 2026 significantly further toward AI automation. For marketers, two developments are especially relevant: Advantage+ Creative takes over more video and translation work, while Meta is also reorganizing attribution for social ads so that reporting in Ads Manager and third-party tools fits together better.
A practical article from Social Media Examiner dated 25 June bundles several of these Facebook Ads changes. The most important product developments can be verified directly against official Meta announcements.
Meta tests UGC videos with AI avatars
Meta is expanding Advantage+ Creative with generative video features. The company is testing, among other things, UGC-style videos with avatars and AI voiceovers, additional voiceovers for existing images and videos, and translations for spoken content and text overlays.
Meta positions these features primarily as a scaling tool: brands should be able to produce more variants for different audiences, markets and formats without manually shooting every version. In a beta test published by Meta, campaigns that used AI video generation in a large share of their ads achieved on average 10 percent higher click-through rates and 8 percent higher conversion rates. These are platform figures and not a guarantee for individual campaigns.
Official source: Meta on AI creatives, creator ads and Reels
For teams, this mainly means the scarce resource shifts from pure production more toward ideas, selection and quality control. This development is also visible more broadly in the market. Our overview AI in media buying: why platform automation needs control explains why automated recommendations and creative production should still be checked independently.
Meta separates link clicks and social engagement more cleanly
On 3 March 2026, Meta announced a narrower definition of click-through attribution. For website and in-store conversions, only genuine link clicks should count in this category. Actions such as likes, shares or saves will instead be assigned to the separate engagement attribution.
At the same time, Meta renamed the former engaged-view attribution to engage-through attribution. For video ads, the company shortened the threshold for an engaged view from ten to five seconds. According to Meta, the rollout should happen gradually starting in March, so it may have become visible at different times depending on the ad account.
The goal: figures from Meta Ads Manager should be easier to compare with systems such as Google Analytics, without removing social interactions from measurement altogether.
Official source: Meta on the new attribution for social ads
For advertisers, this matters more than mere reporting cosmetics. If you compare campaigns across the transition period, you may see shifts in the reported conversion paths even though real demand has not changed to the same degree. For more reliable impact measurement, incrementality therefore remains central. Our article DoorDash has incrementality measurement independently certified fits here.
The Ads Manager also gets more AI guidance
Meta is also testing an AI business assistant for advertisers intended to help with optimization and account support. In January 2026, Meta announced it would gradually expand the assistant and tailor it more closely to the goals and performance of an individual business.
Official source: Meta: 2026 – AI Drives Performance
This fits a broader pattern: TikTok and other advertising platforms are also increasingly shifting campaign work into agentic or heavily automated interfaces. Our article TikTok Ads 2026: AI takes over more and more campaign work offers a direct comparison.
What marketers should practically check now
The most important point is not to switch on every new AI feature immediately. It makes more sense to test three levels separately: creative quality, platform optimization and measurement. A campaign can perform better through more variants and still become harder to interpret in reporting. Conversely, stricter attribution may initially look worse even though it is simply counting differently.
That is exactly why Meta automation should not be measured only by CPA or ROAS in its own dashboard. External analytics, conversion lift tests and clear creative experiments remain important. That many marketing teams use AI intensively but cannot yet cleanly prove the economic effect is also shown by the current study in AI in marketing 2026: usage grows, measurement lags behind.
Source of the feed item: Social Media Examiner: Facebook Ads – New Tools for Better Tracking, More Creative, and Faster Sales