The cockpit controls are gone.
Meta removed detailed targeting interests on August 21, 2025. Any ad set still running deprecated interests stopped cold on January 15, 2026. Google retired modified broad match and is pushing AI Max for Search, which removes keyword targeting altogether.
This is not a platform test. It is a structural shift that is now complete.
What Replaced the Controls
Meta’s targeting engine is called Andromeda. It reads creative elements in real time and uses them to match impressions. The imagery you choose, the copy you write, the hook you open with: those are now the audience signals.
Meta confirmed this directly: “The focus has shifted from niche targeting to creative diversification as the best lever to find the most relevant audiences.”
A fitness product image attracts fitness-interested users. A B2B headline self-selects decision-makers. The creative does the work the interest category used to do.
The lever you are actually pulling is creative supply, not platform settings.
What the Data Says About That Lever
Nielsen found that strong creative drove 86 percent of sales lift in digital ads. AppsFlyer put Meta ad performance attributable to creative quality at 70 to 80 percent in 2025.
The CPM numbers show the same split. Average CPMs across social platforms rose 8 to 12 percent year-over-year in 2025. Well-performing creatives with conversion rates above 10 to 15 percent held CPMs near $25. Weak creatives pushed CPMs past $50.
Same auction. Same platform. Different creative supply. Different cost.
Volume Is Not Optional
Brands that held or improved ROAS in 2025 ran 10 to 20 new ad variations per week. Practitioners running Advantage+ Audience mechanics are pushing 20 to 50 variations per concept.
The structural reason: creative diversity expands the audience segments the algorithm can explore. More creative variety means more signal. More signal means the algorithm finds the right people faster and at lower cost.
Where paid teams used to manage 3 to 5 ads per ad set, that number is now a floor, not a ceiling.
The Trap Inside the Volume Thesis
Volume without quality is noise with a higher production rate.
83.5 percent of marketers say creative drives performance. Only 58.5 percent of those teams share a clear definition of what great creative means, per eMarketer and TripleLift research.
54 percent of marketers admit their assets are not updated often enough, which leads to fatigue.
The volume thesis holds only when the brief is right. An AI Marketing Agent running creative production at scale solves the supply problem. It does not solve the brief problem. Those are two different gaps, and teams that conflate them burn budget on fast production of the wrong angles.
What This Means at the $500K to $10M Scale
Most mid-market paid media teams understand the argument. The gap is operational.
The briefing cadence, the production pipeline, the handoff between strategy and creative: none of that was built for 20-plus variations per week. It was built for a world where the media buyer tuned the targeting and the creative team turned around one or two concepts per campaign.
That world closed on January 15, 2026.
The teams winning now rebuilt the supply chain. They deploy an AI Marketing Agent on the production side so creative velocity matches what the algorithm requires. Claxton Law Group runs an AI Case Acquisition Agent 24/7 and added more than 100K per month in revenue. The same deployment model applies to paid creative supply: the agent runs continuously, the team sets the brief, and volume is no longer the bottleneck.
The best media buyers right now are creative strategists who operate platforms. The platform settings are largely set by the algorithm. The brief and the supply are yours.
The Single Operational Question
Can your creative pipeline produce the volume the algorithm now requires without degrading the brief quality that makes volume worth anything?
If the answer is no, the CPM line will keep climbing regardless of budget.
The teams solving this are not spending more. They are producing more, faster, from a brief that does not change every time a new variation ships.