Global consumer spending via AI agents to hit US$3.35tr by 2030

The next audience marketers need to persuade may not be human. As AI agents increasingly research, compare and purchase products on behalf of consumers, brands will need to rethink how they influence purchasing decisions, according to new research from WARC and PHD.

The report, "From abundance to agents – how the delegation of choice is transforming marketing", argues that marketers are entering a new era where success depends not only on appealing to people, but also to the AI systems that increasingly determine which brands consumers see, compare and ultimately buy.

While consumers will continue making most purchasing decisions themselves over the next few years, AI agents are expected to play a growing role in handling repetitive, information-heavy and comparison-led tasks across the customer journey.

This shift is expected to be significant. The report projects that global consumer spending facilitated by agentic AI will more than triple from US$944 billion in 2026 to US$3.35 trillion by 2030, accounting for 3.8% of global consumer spending, up from 1.3% this year.

Rather than viewing AI as simply another marketing channel, the report argues that brands should see AI agents as a new decision-making layer between consumers and businesses. This means marketers will increasingly need to optimise for both human audiences and machine intermediaries. While creativity, storytelling and emotional connections remain essential for building brand equity, brands will also need to ensure their products, services and brand signals are understandable to AI systems responsible for recommending purchases.

Don't miss: How can brands win hearts when AI does the shopping?

Where AI will make the biggest impact

The report identifies telecoms and utilities, financial services, and travel and transport as the industries most likely to transition towards agent-to-agent interactions, where AI handles much of the comparison, recommendation and purchasing process.

Telecoms and utilities are forecast to become the largest category for agent-facilitated transactions by 2030, with spending expected to surge more than 600% from US$57.6 billion in 2026 to US$410.3 billion. Frequent billing cycles, comparison-heavy purchasing journeys and information-rich products make the sector particularly suited to AI-led decision-making.

Financial services is expected to see agent-facilitated consumer spending increase by 235.3% to US$237.9 billion by 2030. Rather than replacing consumers, AI is expected to increasingly narrow down mortgage, insurance and banking options before people make the final decision.

Meanwhile, travel and transport is projected to grow from US$78.1 billion to US$275.6 billion over the same period as AI agents increasingly manage itinerary planning, price comparisons and bookings.

Beyond these sectors, media and publishing, food, retail and soft drinks are also expected to experience rapid growth as consumers increasingly rely on AI for recommendations and repeat purchases.

From emotional storytelling to machine-readable marketing

To help marketers understand where AI will have the greatest influence, PHD introduced its "Four modes framework", which categorises industries based on whether purchasing decisions are primarily driven by brands, consumers or AI agents.

The framework outlines four modes of influence: Brand → Consumer, Consumer → Consumer, Agent → Consumer and Agent → Agent. Rather than suggesting AI will replace traditional marketing, it argues that all four will coexist, with their influence varying by category, purchase occasion and stage of the customer journey.

For marketers, the shift is less about replacing creativity and more about expanding what optimisation looks like.

While emotional storytelling, memorable campaigns and distinctive brand assets will remain critical in building awareness and preference among consumers, brands will also need to ensure AI systems can accurately understand and recommend their products and services.

In practice, this means investing in structured, machine-readable product information, maintaining consistent brand assets across digital touchpoints, and communicating clear value propositions that AI agents can interpret when comparing options.

The report also suggests that trust signals, including customer reviews, transparent pricing, service quality and brand reputation, could increasingly influence whether an AI system recommends one brand over another.

As AI agents become a more prominent part of the customer journey, marketers may need to optimise not just for search engines or social algorithms, but also for the AI systems that increasingly shape product discovery and purchase decisions. In that environment, winning consumer attention may no longer be enough; brands will also need to earn the recommendation of the machines acting on consumers' behalf.

Related articles: 
IMDA sets guardrails for agentic AI with new framework  
Why SEO alone won't win AI recommendations  
Showing up in AI answers isn't enough if audiences don't believe them