Search Here

The gap between expectations and outcomes surrounding artificial intelligence in the agency sector reveals a notable divergence: anticipated disruption has proven less severe than predicted, while positive impacts have exceeded expectations. Research surveying 181 agencies in September 2025 found that 34% expected negative revenue impact from AI adoption.

However, only 27% actually experienced negative effects. Meanwhile, 65% report positive revenue or profit impact, suggesting the technology's benefits have been significantly underestimated relative to its risks.

This pattern, overestimating threats and underestimating opportunities, appears across multiple dimensions of AI's agency impact.

Regarding pricing pressure, 47% of agencies expected clients to request rate reductions due to AI implementation. In reality, only 27% received such requests, and merely 13% actually lowered prices. The feared race to bottom pricing has largely failed to materialize.

On employment, widespread predictions of AI-driven layoffs have not manifested. Only 3% of agencies made significant staff reductions specifically attributable to AI, while 64% have no layoff plans. Though 21% are considering reductions, the overwhelming majority of agencies have maintained stable employment despite substantial AI adoption.

"We've embedded AI across all our teams," said Tom Schofield, Operations Director at Engage Interactive. "It's lifted efficiency across the board and even reduced our dependency on external services like translation."

The positive revenue impact agencies report stems primarily from operational optimization rather than revolutionary new capabilities. Common improvements include automating administrative tasks, streamlining sales pipelines, accelerating production workflows, and scaling output without proportional hiring.

Jake Skoric, COO of Delta Reality, described his firm's experience: "Our 3D design team integrated AI tools into the production pipeline, which increased efficiency by three to four times. This has reduced utilization but allows us to handle significantly more work with the same team size."

This represents a pattern of incremental improvement accumulating into significant business impact, a quieter transformation than dramatic disruption.

However, the data also reveals where expectations remain unfulfilled. While 65% report positive operational impact, only 13% have created meaningful new revenue streams from AI-specific services. The gap between internal optimization and external innovation suggests agencies are successfully defending existing business but have yet to pioneer next-generation offerings at scale.

Industry observers note that the expectation-reality gap may reflect the AI narrative's dominance by extreme positions. Media coverage has oscillated between utopian promises of exponential growth and dystopian warnings of widespread unemployment, with limited attention to incremental, practical applications.

"I think many companies are playing 'wait and see,'" observed Thomas Van Sundert, co-founder of November Five. "With any revolution, the world repositions itself and it opens up opportunities. Companies still don't know where it'll land and I think that's why they're holding off on making any harsh decisions."

This cautious pragmatism appears to be serving agencies well. Rather than wholesale transformation, most are iteratively integrating AI into existing workflows, measuring results, and adjusting approaches, a methodology more aligned with continuous improvement than revolutionary change.

The research methodology may influence results. Survey participants were customers of agency management platform Productive, potentially skewing toward digitally mature, process-oriented firms typically in the 20–50 employee range. These characteristics might predispose organizations toward successful AI integration.

Nevertheless, the data suggests that mid-market agencies which are often considered most vulnerable to AI disruption due to performing routine tasks potentially subject to automation are instead finding practical applications that enhance rather than replace their capabilities.

The divergence between expectations and outcomes carries implications for strategic planning. Agencies that prepared primarily for defensive scenarios, protecting against job losses, pricing pressure, and market commoditization, may have underinvested in offensive strategies like developing new AI-enabled services or premium positioning based on AI-augmented quality.

As one agency operations manager noted, the real shift is less about AI replacing human work and more about redefining what constitutes value: "AI helps us work faster without compromising quality."

The transition from panic to profit appears less about AI's capabilities fundamentally changing agency work and more about agencies discovering that augmentation creates different opportunities than replacement.