The Agency of Record Is Being Rewritten: What Revenue Growth Means for Challenger Brands in an AI Era

Forrester’s Predictions 2026 report on marketing agencies makes an unusually blunt call: agencies, as an industry, are about to be materially changed by the same AI wave they’ve spent two years selling to clients. The report points to low-margin, project-based work replacing what used to be reliable retainer fees, and names AI and automation directly as forces disrupting agencies’ labor-based economic model.

That’s not a prediction about a few agencies losing accounts. It’s a prediction about the model itself.

Why agencies can’t out-invest in AI, structurally

The reason isn’t a lack of talent or ambition inside agencies. It’s arithmetic. An agency’s revenue is built on billable hours: the more efficient AI makes a given task, the fewer hours there are to bill for it. Every dollar an agency invests in automation that removes hours from a retainer is a dollar working against its own revenue line.

That’s the same pattern Clayton Christensen documented in disk drives and steel minimills: incumbents don’t lose to disruptive technology because they’re incompetent; they rationally underinvest in the thing that would undercut their own current revenue. It’s playing out in marketing services now, and the data already shows it.

What the numbers say

Gartner’s CMO Spend Survey has been tracking the shift for two consecutive years. In the 2025 survey, 39% of CMOs said they planned to cut agency budgets, largely by eliminating unproductive relationships and streamlining rosters, and 22% specifically credited generative AI with reducing their reliance on external agencies for creative and strategy work. By the 2026 survey, that shift had become a funding source: CMOs pushed paid media allocation up to 31.4% of budget, funded in part by continued cuts to agency spend.

Put simply, the money that used to sit in a retainer line is moving into media and, increasingly, into the AI-powered revenue growth platform layer that lets a smaller in-house team do what the agency used to do.

The Agent of Record: a different economic model, not just a different vendor

This is the distinction behind what Insika AI calls the Agent of Record, a deliberate rework of the industry’s Agency of Record. The difference isn’t cosmetic:

An agency’s best strategist still has fifteen other clients. A decisioning engine built as your Agent of Record doesn’t.

What this looks like for a challenger brand

This matters most for brands that can’t outspend the category leader and have to out-execute instead. GURU Organic Energy, a challenger brand in a crowded energy-drink category, used Insika’s go-to-market AI to cut planning time by 90% and unlock the digital spending that correlates back to increase in-store sales. That’s the practical shape of AI for challenger brands: it’s not about having a bigger budget than the incumbent, it’s about closing the distance between a market signal and a shipped campaign from weeks to days.

For a $300M-$1B consumer or B2B brand carrying multiple agency relationships and a lean internal team, that speed differential is often the entire competitive edge available to it.

Where agencies still fit

None of this means every agency relationship is the enemy. Boutique creative shops, production partners, and specialist execution teams solve a real and different problem than the retainer-based agency-of-record arrangement that owns a brand’s growth mandate. The disruption argument is aimed squarely at that mandate, the one built on hours and quarterly strategy decks, not at the creative partners who do original work inside a faster system.

The practical distinction: if an agency’s value is billed hours managing your channels, that’s the relationship an AI revenue growth platform is built to absorb. If its value is a creative idea you couldn’t produce internally, that’s a different conversation entirely.

Sources: Forrester, Predictions 2026: Marketing Agencies Resign Their Agency (October 2025); Gartner, 2025 and 2026 CMO Spend Surveys; Insika AI customer results.