The Pitch Has Changed: How AI Is Rewriting Agency New Business

From simulated founder interviews to predictive creative testing, the agencies winning accounts are the ones that show up with tools their competitors don't have

When Noble People walked into a pitch for a tech company last year, they opened by telling the client they’d already spoken to the founders. That wasn’t exactly true. The agency had trained a custom GPT on hundreds of the founders’ public interviews to simulate their voices and test ideas through that lens. When they revealed the trick, the room shifted.

“The vibe in the room changed,” said Tom Morrissy, the agency’s chief growth officer. “I knew we had them, no matter what happened after that.”

This is what agency pitches look like now. The technology is helping agencies move faster through typically time-consuming and costly reviews. And for the holding companies that just completed the largest agency merger in history, AI isn’t just a service offering—it’s the justification for the deal itself.

The Pitch Theater of 2025

When Omnicom’s executives were asked during their Q3 earnings call how AI is reshaping creative and media, the chief technology officer went straight to new business pitches. He pointed to a recent win for a large automotive company where “integrated agents” helped guide consumer research, creative concepting, and production.

The pattern is consistent across the industry. IPG’s AI-powered Interact platform helped win Bayer’s global consumer health review by uncovering repositioning insights for the Canesten brand—analyzing campaigns from beauty and luxury categories to “break out of consumer health thinking.” Using a partnership with AI startup Aaru, IPG tested creative concepts and predicted how consumers might respond to them weeks later in retail settings.

Horizon Media built a custom analysis inside its Blu platform for its Spectrum pitch, combining the telecom company’s public market data with the agency’s own spine of 260 million U.S. consumer profiles. Teams could ask questions like “Which audiences are most likely to churn in Los Angeles?” and receive insights that previously would have taken weeks.

Fig used its proprietary Story Data platform to win Tropicana’s creative consolidation, mapping publicly available creative assets across the juice category to show how formulaic the category had become—and making the case for something different.

What Changed

The speed advantage is obvious. Agencies can now do in hours what used to take weeks. But three deeper shifts are reshaping the competitive dynamics.

The tool itself becomes the differentiator. When agencies build proprietary AI capabilities, they create what one strategist called “a value moat.” Clients know they can’t DIY what’s being offered. Independent agencies, often more agile than holding company networks, are capitalizing by building custom tools rather than licensing generic platforms.

Predictions replace recommendations. The agencies winning pitches aren’t just offering strategies—they’re offering predictions. “That’s amazing because that’s a predictor of behavior, not just a predictor of response,” Bayer said about IPG’s approach. The shift from “here’s what we recommend” to “here’s what will happen” changes the client conversation entirely.

Speed becomes proof of capability. Bayer’s pitch included briefs where agencies were expected to deliver creative work by end of day. The ability to produce that quickly—with AI assistance—became part of the evaluation criteria.

The Holding Company Calculation

The Omnicom-IPG merger—valued at roughly $8.9 billion—is explicitly framed around AI and data capabilities. The combined entity expects $750 million in annual cost savings while creating a revenue base that rivals Accenture Song. Omnicom leadership has cited precision marketing and data-driven capabilities as the biggest areas of synergy.

But the merger also reflects defensive pressure. Digital platforms like Google, Meta, and Amazon increasingly enable brands to bypass agencies entirely, controlling over 50% of global digital ad spending. Consultancies like Accenture challenge agencies by offering integrated solutions that combine strategy, technology, and execution. IPG’s CEO initiated strategic talks after recognizing that creative strength alone couldn’t offset eroding client budgets.

The question for clients: does bigger actually mean better? Industry commentary suggests skepticism. “The brands are the storefront, because clients still like to hire brands,” said one agency president, “but there will be little customization, more principal buying within media, and continued resource pressure.” Integration adds bureaucracy, complexity, and delays that can undermine the responsiveness modern marketing demands.

The Independent Opportunity

As holding companies consolidate, independent and challenger networks are positioning differently. Stagwell has set a goal to double revenue to $5 billion by 2029 through tech-driven acquisitions—its Marketing Cloud division posted 31% year-over-year growth. Attivo has acquired established agencies that larger holding companies could no longer efficiently manage, including Hill Holliday and Deutsch NY from IPG.

The pitch to clients: we’re faster, more flexible, and more willing to build custom solutions rather than forcing you into scaled platforms.

For AI-specialized agencies, valuations are climbing. Agencies with expertise in the industry’s current favorite technology are expecting significant interest. But the pressure cuts both ways. As one industry observer noted, “If last year’s deals were any clue, agency M&A will stay hot—driven by performance, data, and tech integration.”

What Clients Should Ask

If you’re evaluating agencies, the AI questions have changed from “do you use AI?” to more specific probes:

What’s proprietary versus licensed? Agencies using the same off-the-shelf tools as everyone else can’t claim differentiation. Ask what they’ve built themselves.

How does AI affect pricing? If AI dramatically reduces production time, that should show up somewhere—either in lower costs or more deliverables.

What predictions can you make? The best AI applications aren’t just faster—they’re predictive. Ask for examples of predictions that proved accurate.

How do you handle creative homogenization? When everyone pulls from the same data and feeds similar prompts into similar tools, output converges. What’s the approach to maintaining distinctiveness?

The pitch has changed. The agencies winning are the ones that show up with tools their competitors don’t have—and predictions their competitors can’t make.