The Extinction Level Event

I'm often asked about the impact of AI in agencies, with enquirers interested in automated media trading, dynamically personalised content and new workflows and the like.

My answer is the same: The impact of AI in agencies pales in comparison to the impact of AI upon agencies.

More specifically how AI is powering the potential "ELE" (Extinction Level Event) in the form of the platforms disintermediating the agency layer in the marketing value chain by working directly with clients.

A decade ago, Facebook (as Meta was then) tried to establish a creative services unit that could work directly with clients. They learned quickly that this was more complex and human than they expected and stepped back from this and renewed the frenemy relationship with agencies.

This time is different. It's still complex, but less human. And the complexity can be manage by automation. Hence the ELE.

What I shared six months ago is happening faster than even I predicted.

In March, I shared my analysis of how Meta's Advantage+ and Google's P-Max were testing the waters for enterprise-level agency disintermediation. Several of you wrote back noting similar patterns, with increased pressure to work directly with platforms, agencies struggling to justify their strategic value, and performance metrics that increasingly favoured algorithmic optimisation over human expertise.

What I didn't anticipate was the velocity. This week's announcement from TikTok and developments recently from Amazon don't just confirm the trend, they complete a strategic encirclement that leaves traditional agency models with no viable escape routes.

The timing of recent platform announcements reveals strategic coordination rather than independent innovation. Meta established the template with Advantage+, Google expanded it with P-Max, and now TikTok and Amazon have completed what can only be described as a coordinated assault on traditional agency functions.

Meta's Foundation: Advantage+ Shopping didn't just automate media buying—it demonstrated that algorithmic optimisation could consistently outperform human strategic decision-making. Businesses using Advantage+ saw up to 32% improvement in return on ad spend, creating a compelling narrative that surrendering control delivers superior results.

Google's Amplification: P-Max expanded this approach across search, display, video, and shopping while introducing conversational AI agents that provide 24/7 optimisation recommendations. Their Marketing Advisor now learns from performance data across millions of campaigns, offering strategic insights that dwarf any single agency's experience base.

TikTok's Entertainment-Commerce Integration: TikTok's new AI tools represent the most sophisticated approach yet—embedding commerce directly within entertainment consumption. Rather than interrupting content with advertising, they've created closed-loop systems where discovery, consideration, and purchase happen within a single platform experience.

Amazon's Commerce-First Advantage: Amazon's retail media evolution combines first-party purchase data with AI-powered creative optimisation, eliminating traditional media planning by automatically adjusting spend based on real-time conversion patterns. They offer brands direct access to consumers at the exact moment of purchase intent.

The strategic coordination becomes undeniable when viewed collectively. These platforms aren't competing with agencies, they're systematically replacing every core agency function while offering superior data access, measurement accuracy, and conversion performance.

While Meta, Google, TikTok, and Amazon attack from media and commerce angles, Adobe approaches from customer experience orchestration, arguably the most sophisticated disintermediation strategy because it appears collaborative while systematically transferring strategic control.

Adobe's Brand Concierge and Experience Platform Agent Orchestrator don't just automate campaign management, they assume control of the entire customer journey from awareness through retention. When Adobe CEO Shantanu Narayen received Cannes Lions' first "Creative Champion" award for "democratising creativity through technology," it signaled industry validation of platform-controlled creative processes.

The orchestration approach is particularly insidious because it positions itself as empowering brands while gradually transferring strategic authority to algorithmic decision-making. Brands maintain the illusion of control while Adobe's systems learn their preferences, optimise their processes, and eventually anticipate their needs better than internal teams.

This creates complete encirclement: platforms now control media buying, creative production, audience targeting, performance optimisation, customer experience orchestration, and strategic recommendation. The remaining agency functions such as relationship management and industry expertise become increasingly difficult to defend when clients achieve measurable superior results through direct platform relationships.

Over the past six months, I've heard agency leaders articulate two primary defense strategies. Both are failing under algorithmic competition that operates faster, with more comprehensive data, and delivers measurable superior results.

The Agency Argument: "We provide strategic guidance across all platforms, optimising your investment portfolio independent of any single platform's commercial interests."

The Algorithmic Reality: AI-powered automated budget allocation now occurs in real-time based on performance data rather than human strategic assumptions. Google's enhanced Data Manager connects first-party data sources including BigQuery, HubSpot, Oracle, Salesforce, and Shopify, creating cross-platform optimisation that surpasses human strategic decision-making.

When Amazon's AI can automatically shift budget from Meta to TikTok based on conversion performance within minutes, the value of quarterly strategic planning sessions diminishes dramatically. Platforms argue—with performance data to support their claim—that algorithmic optimisation provides more responsive, data-driven investment decisions than human strategists constrained by monthly reporting cycles.

The Agency Argument: "We help you own and control your customer data, preventing platform dependency while building valuable first-party data assets."

The Platform Counter-Strategy: Platforms offer increasingly sophisticated first-party data integration while maintaining their closed ecosystem advantages. Meta's advanced AI translation features, Google's Chrome-integrated Marketing Advisor, and Amazon's retail media insights provide data value that agencies cannot replicate.

Clean-room technologies, server-side tracking, and value-for-data exchanges enable brands to maintain data control while accessing platform optimisation capabilities. Platforms position this as "data independence with platform performance"—having strategic control without sacrificing algorithmic advantages.

The uncomfortable truth several Brandflow subscribers have shared privately: customer data ownership becomes less strategically valuable when platform ecosystems deliver superior performance outcomes. The question shifts from "Who controls our data?" to "How do we achieve competitive advantage using platform capabilities while maintaining strategic independence?"

My analysis earlier this year identified agencies caught between integrated giants and specialists as particularly vulnerable, what I called "The Deathly Middle." Platform disintermediation has accelerated this vulnerability into existential crisis by attacking the core functions that sustained middle-tier agencies:

Media Planning and Buying: 85% of Meta's ad revenue now flows through automated buying, with Google's AI-powered tools consistently outperforming human alternatives in head-to-head testing. Middle-tier agencies built on media expertise face direct algorithmic competition.

Creative Optimisation: AI-generated headlines, imagery, and video creative now outperform traditional agency creative in performance metrics. Platforms provide creative tools that deliver measurable superior results while reducing production costs by 40-60%.

Account Management: Conversational AI agents provide 24/7 optimisation recommendations, proactive problem-solving, and strategic insights drawn from performance data across millions of campaigns. Human account management struggles to compete with this scale and responsiveness.

Strategic Planning: Platform algorithms optimise campaigns at millisecond speeds across millions of variables, making human strategic planning appear slow and limited by comparison.

The timeline for strategic response has compressed from years to months. For agencies in the "Deathly Middle," three paths remain viable, and the window for choosing is narrowing rapidly:

The resistance to platform disintermediation often stems from organisational inertia rather than strategic analysis. When AI tools deliver 90% of traditional agency value at 20% of the cost, the economic logic becomes overwhelming regardless of relationship preferences or historical arrangements.

The platforms understand this dynamic and structure their enterprise expansion accordingly. They begin with performance-focused pilot programs that demonstrate measurable superior results, gradually expanding scope as clients become comfortable with algorithmic decision-making.

Marketing leaders who wait for perfect platform solutions or hope for agency competitive responses risk falling behind competitors who embrace platform-direct strategies earlier in the adoption cycle.

The agency disintermediation accelerator doesn't offer comfortable choices. It demands strategic clarity about which battles can be won and which evolutionary paths lead to sustainable competitive advantage.

The platforms are building toward complete marketing ecosystem ownership. The question for marketing leaders isn't whether this transformation will occur, but whether their teams will lead this transition or struggle to adapt when platform-direct becomes the competitive baseline rather than the innovative exception.

Brandflow is written by Justin Billingsley, who has spent his career on all three sides of the industry's table: senior client, global agency leader, technology founder. First published 22 August 2025 in the Brandflow newsletter on LinkedIn.