Google Ads Expands AI Max Testing as Firms Weigh AI ROI

A New Layer of Testing for Google Ads Campaigns

Google Ads has rolled out expanded experimentation tools for Search and its AI Max campaign features, giving advertisers more structured ways to test budget levels, ROI targets, brand safety controls, location settings, and other campaign changes before committing spend broadly 1. The update reflects a broader shift in digital advertising toward built-in experimentation, letting marketers validate performance assumptions rather than relying on guesswork or post-hoc analysis. For an industry increasingly leaning on AI-driven bidding and targeting, the ability to run controlled tests against defined ROI benchmarks marks an incremental but meaningful step toward more accountable AI-powered marketing spend 1.

The Bigger Question: Does Enterprise AI Actually Pay Off

The timing of Google's move lands amid a wider reckoning across industries about whether AI investments are delivering measurable returns. At The Six Five Summit: AI Unleashed 2026, Intel's Anil Nanduri argued that enterprises evaluating AI infrastructure should prioritize return on investment over raw hardware specifications, suggesting that many organizations are still anchoring purchasing decisions to processing power rather than business outcomes 2. His comments point to a maturing conversation in enterprise AI adoption, where the emphasis is shifting from capability for its own sake to demonstrable financial or operational impact.

That shift is proving harder in some sectors than others. In healthcare, Forbes reports that traditional ROI frameworks may not adequately capture the value AI tools provide, since benefits can show up in less quantifiable ways—improved diagnostic confidence, reduced clinician burnout, or better patient triage—rather than straightforward cost savings 3. This suggests that as AI spreads into specialized industries, measurement standards developed for one domain, such as advertising, may not translate cleanly to another, such as clinical care.

Mixed Signals From the Corporate World

Adding to the uncertainty, Fortune highlights findings buried in OpenAI's own research showing no clear correlation between AI use and revenue per employee among corporate ChatGPT customers, a striking admission given the company's stake in proving enterprise value 5. This complicates the narrative that AI adoption straightforwardly boosts productivity or profitability, reinforcing calls like Nanduri's for more rigorous, outcome-based evaluation rather than assumptions tied to adoption rates or compute power 25.

Skills and Strategy Catching Up

Meanwhile, the workforce implications of this shift are becoming part of professional training itself. Coverage of digital marketing education notes that courses now incorporate real business case studies, including material from Harvard, to help marketers think more strategically about how AI tools fit into decision-making 4. Together, these threads—new experimentation infrastructure from Google, skepticism from hardware and research leaders, sector-specific ROI challenges in healthcare, and evolving marketing education—illustrate an industry still working out how to prove, rather than assume, that AI investment translates into real returns.