Meta Shows What AI-Scaled Advertising Efficiency Looks Like
Meta’s latest results highlight how AI ad optimization is turning a largely fixed audience into a far more profitable one. The company reported revenue of USD 56.31 billion (approx. RM264.0 billion) for the quarter, a 33% year-on-year jump, driven mostly by advertising. Yet daily active users grew only 4%, signaling that the real story is efficiency, not reach. Meta’s AI-powered ad engine is tuning every layer of programmatic advertising—from targeting and campaign optimisation to pricing and inventory utilisation. Ad impressions climbed 19% while average pricing increased 12%, allowing Meta to serve more ads and charge more for them without expanding its user base dramatically. For advertisers, this means better performance potential but also higher dependence on Meta’s automated ecosystem, where AI bidding strategies and delivery decisions are increasingly controlled by the platform rather than human media buyers.
Creative Performance Analysis: From Guesswork to Fingerprints
While platforms like Meta optimise delivery, a new wave of tools is transforming creative performance analysis itself. Monks’ Creative Intelligence engine, part of its Monks.Flow agentic AI system, is designed to decode exactly why certain ads work. It ingests video and static assets, automatically segments them into clips and generates detailed metadata on elements such as expressions, product close-ups or background tracks. These creative “fingerprints” are then stitched directly into media performance metrics, revealing which precise sensory moments drive engagement and conversions. Backed by findings that creative accounts for nearly half of sales impact, this approach reframes creative as the new targeting layer in programmatic advertising. Instead of treating assets as static units, marketers can scale winning moments across formats and audiences, accelerating testing cycles while reducing wasted spend on underperforming variants.
Self-Serve AI Ad Managers Are Democratizing Optimization
AI-powered campaign optimization is no longer confined to the biggest walled gardens. OpenAI’s beta Ads Manager for ChatGPT signals how AI ad optimization is becoming accessible to a wider range of advertisers. Through a self-serve portal, marketers can set budgets, upload ads, define pacing and monitor performance in a familiar dashboard. The introduction of cost-per-click bidding aligns spend directly with user actions, strengthening AI bidding strategies that optimise for outcomes, not just impressions. OpenAI is working with major agency holding companies and technology partners to integrate ChatGPT ads into existing workflows, helping brands plug AI-driven automated ad management into their current media stacks. With OpenAI expecting to generate USD 2.5 billion (approx. RM11.7 billion) in ad revenue this year and eyeing far larger numbers by 2030, the company is positioning its platform as a new locus of performance media beyond traditional social networks.
Agentic AI Systems and the Next Wave of Automated Ad Management
Beyond isolated tools, agentic AI systems are emerging as the next frontier in automated ad management. Platforms like Monks.Flow and Taboola’s agentic AI solutions aim to orchestrate entire marketing workflows—planning, execution and optimisation—through interconnected AI agents. Taboola’s recent study shows that 76% of advertisers already see meaningful performance uplift from AI-powered solutions, primarily on search and social. Yet many feel boxed into these environments. A striking 80% say they would increase spend on the open web if comparable agentic capabilities were available, and most are willing to reallocate a significant portion of their performance budgets. The challenge is integration: larger advertisers, especially those spending in the millions monthly, report that embedding agentic AI into complex legacy workflows is a dominant barrier. As these systems mature, they promise always-on optimisation that continuously predicts, tests and rebalances campaigns across channels without human micromanagement.
The New Dependency: Value and Risk in AI-Driven Advertising
As AI bidding strategies, creative intelligence tools and agentic systems become standard, advertisers face a strategic paradox. Platforms are clearly extracting more value from existing audiences, as Meta’s AI-driven revenue lift demonstrates, and tools from OpenAI, Monks and Taboola show how programmatic advertising can be optimised end-to-end. But the more performance depends on proprietary AI, the more leverage shifts toward platform owners. Efficiency gains do not automatically translate into lower costs for advertisers; they often bolster platform pricing power instead. Meanwhile, workflow integration challenges limit how quickly large brands can diversify beyond dominant ecosystems. The next phase of automated ad management will likely revolve around negotiating this dependence: balancing the upside of AI ad optimization and automated ad management with the need for flexibility, cross-channel measurement and control over the data and models that increasingly determine campaign outcomes.