From Gut Feel to Creative Performance Prediction
Marketing teams are moving away from intuition-led creative reviews and towards AI ad optimization that forecasts outcomes before human sign-off. Monks’ Creative Intelligence engine illustrates this shift. Embedded in its Monks.Flow ecosystem, the system ingests video and static ads, automatically segments them into clips, and generates detailed metadata on elements such as facial expressions, product framing, and audio cues. Those data points are then stitched directly to media performance metrics, turning creative performance prediction into a repeatable process rather than a guessing game. Jakub Otrzasek, SVP of data at Monks, describes it as isolating the exact second or image that drives conversions, such as a brief shot of a product in motion. For marketers, the implication is that creative is no longer a static asset but a dynamic variable that can be optimized and scaled long before the creative director’s final review.
Platform AI Engines Are Rewriting the ROI Equation
Meta’s latest results show how platform-owned AI is changing what drives advertising growth. With revenue reaching USD 56.31 billion (approx. RM264.0 billion) and advertising contributing USD 55.02 billion (approx. RM258.2 billion), the company is extracting more value from its existing users rather than depending on explosive audience expansion. AI now touches ad targeting, pricing, inventory use, and creative recommendations, enabling Meta to serve more ads and increase average pricing simultaneously. For marketers, this means AI ad optimization can improve campaign outcomes, but the efficiency gains are often captured by the platform through stronger pricing power. As Meta invests up to USD 145 billion (approx. RM679.0 billion) in AI-related infrastructure and data centres, its ecosystem becomes harder to replicate. That deepens advertiser reliance on platform-native programmatic advertising tools, shifting leverage away from brands and toward the algorithms that manage their spend.
Agentic AI and the Automation of Creative Work
The rise of agentic AI is extending automation beyond bidding and targeting into the creative layer itself. Monks’ Creative Intelligence engine functions like a creative co-pilot: it fingerprints assets, identifies the visual and sensory cues that correlate with higher engagement, and feeds those insights back into production. This allows AI-generated ad creatives to be tailored and iterated at scale, with specific scenes, colours, or soundtracks dialed up based on performance signals. Taboola’s research underscores how attractive this model has become. Its study on agentic AI shows that 76 per cent of advertisers already see meaningful performance uplift from AI-powered solutions in search and social. Marketers now expect the same “always-on” optimization for creatives that they get from automated bidding. As these tools mature, parts of storyboarding, versioning, and even concept refinement are increasingly handled by software, reserving human input for higher-level brand and narrative decisions.
Creative Autonomy vs. Ecosystem Dependence
AI-driven optimization is forcing marketers to rethink where strategy ends and automation begins. On one hand, platforms like Meta and engines such as Monks.Flow deliver measurable gains by linking granular creative attributes to real-time performance. On the other, they increase dependence on closed ecosystems that control both data and delivery. Taboola’s study highlights this tension: 80 per cent of advertisers say they would immediately increase spend on the open web if comparable agentic AI solutions existed, and 86 per cent are willing to shift up to a quarter of their performance budgets. Marketers want AI-grade automation without surrendering too much creative autonomy or bargaining power to walled gardens. The emerging playbook is hybrid. Teams lean on platform AI for optimization and scale while using independent tools to decode why specific creative elements work, ensuring that brand voice and long-term positioning still come from human judgment, not just machine recommendations.
Redefining How Creative Performance Is Measured
As AI systems move upstream in the creative process, they are also redefining how success is measured. Instead of assessing campaigns at the end of a flight, marketers can now track performance at the level of frames, scenes, and micro-moments. Monks’ Creative Intelligence links metadata to outcomes so precisely that teams can identify a single 0.04-second clip that lifts conversion, then programmatically replicate its attributes in future assets. Meanwhile, Taboola’s findings that “results are following” AI adoption validate this model: advertisers are not just experimenting with agentic AI, they are seeing quantifiable uplift that justifies budget shifts. Combined with Meta’s AI-enhanced monetisation engine, a new standard is emerging in which creative performance prediction is continuous and predictive, not episodic and retrospective. The creative review meeting is no longer the starting point of performance discussion; it is increasingly the final human checkpoint before AI-optimized campaigns go live.