Digiday+ Research: Agencies punt budget growth expectations to 2027 — while AI worries intensify

This continued the trend of spending pessimism from 2025: Forty-seven percent of agency pros said reduced client budgets was agencies’ top concern last year, ahead of other roadblocks such as scope creep (16%) and the external effects of AI (11%).

However, reduced client budgets isn’t as big of a concern in 2026 as it was in 2025 — down 9 percentage points in 2026 versus 2025. And, some agencies are looking even further ahead and expect greater spending in 2027, based on conversations they’ve had with clients.

Rob Davis, president and CMO at Novus, and Scott Shamberg, president and CEO at Mile Marker, told Digiday last fall that clients have promised bigger budgets in 2027.

“A couple of new business pitches have noted the 2026 budget is X [amount], but the 2027 budget is already projected to be X times, plus 50%,” Davis said. “[The message is] ‘we’re still going to wait and see, but then the next year — 18, 24 months out — is when [we’ll spend].’ It was almost like [hearing], ‘our budgets aren’t that big, but trust us, they’re going up when we think things will really settle, which is the following year.’ I had never seen that before.”

“We’ve seen that too, and I had never seen that before either,” Shamberg added. “I’m interpreting that as, ‘I know the budget for next year [2026] is not exactly what you would want, but the next year [2027], I swear, is going to be huge.’”

Agencies’ concerns over AI’s impact deepen

While client budgets dipped slightly in importance as a concern for agencies in 2026, the effects of AI on the industry took on greater significance. Thirty-eight percent of survey respondents said the biggest challenge the agency industry faces in 2026 is the effects of AI, including the external effects of AI, such as new AI tools (29%). The internal effects of AI, such as workforce development (9%), were also a concern.

By contrast, in 2025, the effects of AI were much less of a concern for agency pros, with only 11% of survey respondents selecting the external effects of AI as a concern last year and none selecting the internal effects of AI. The second biggest problem agencies cited for last year was scope creep, selected by 16% of respondents.

Overall, agency clients lack familiarity with some types of AI. When Digiday asked agencies about their clients’ familiarity with agentic AI for the 2025 Media Agency Report, nearly three-quarters of agency respondents (73%) said their clients do not understand what agentic AI is and how it can be used to optimize ad campaigns.

Agency holding companies are also struggling to find an effective AI business model. According to Digiday reporting on recent hold cos’ earnings calls, companies including WPP, Publicis, Omnicom, Havas and Dentsu are using the same AI language, layers, agentic orchestration and human-supervised automation. As a result, AI strategies don’t sound like differentiated offerings to clients, who say they aren’t seeing a cost savings.

“The honeymoon around AI and agencies is essentially over,” said Gartner vp analyst Jay Wilson. “A year ago, agencies were selling clients on their AI platforms and saying this is going to create cost efficiencies and speed to market. CMOs said, ‘is it going to cost less?’ And agencies were saying, ‘well, not quite yet’. And now, as I talk to CMOs who are reviewing and renewing agency scopes of work, they’re still not seeing the cost savings in terms of fee.”

Agencies and clients are also concerned about the changing dynamics of search marketing and the effects of zero-click search on traditional advertising. Agencies and platforms have been racing to offer clients zero-click analysis tools, from Jellyfish’s Share of Model platform to SEO firms Ahrefs and Semrush, to WPP’s Generative UI venture, built via a partnership with Google.

But, monetizing the tools isn’t easy. For a brand to remedy its zero-click search performance, it needs to improve either brand awareness or earned media exposure, priorities advertisers already constantly work to address. That realization can lead clients to abandon tools after initial trials.

“They basically wanted a secret hack that would suddenly allow them to rank in ChatGPT,” said Benjamin Houy, an app publisher in London who offered brands and agencies a zero-click monitoring tool, Lorelight, for about six months in 2025 before discontinuing it. “At the end of the day, they were doing the same thing they were doing before the [Lorelight] insights. There was no behavior change… Why pay for it?”