An equal number of CIOs (53%) prioritize both productivity gains and revenue growth as their success metrics for AI adoption — and just 42% monitor employee satisfaction, new research revealed. The survey — fielded by revenue intelligence provider Gong — explored how this divergence underscores a broader challenge: Confusion about where AI can deliver the most business value and a well-defined approach for evaluation.
“These competing priorities ultimately stem from AI’s expanding use cases,” said Eilon Reshef, Co-Founder and Chief Product Officer of Gong, in an interview with Demand Gen Report. “As the technology becomes more widely adopted in the enterprise, it has become known primarily as an efficiency tool that can reduce the time spent on administrative tasks. While this is true, as AI capabilities expand, so do the ways companies can implement it to find value. As a result, measuring its value becomes more complex.”
Key insights from the study include:
Revenue Growth Vs. Time Savings
Worldwide, 61% CIOs believed increased revenue alone justifies AI costs, while 60% said that time savings alone will justify costs. Yet only 32% actively measure both, suggesting that many companies still don’t have systems in place to measure and assess the impact on the variables that they say matter most.
Reshef explained that this is likely because measuring revenue and time savings isn’t an “apples-to-apples” comparison, and many CIOs likely lack a decisive way to measure both metrics simultaneously. Instead, he explained that revenue organizations may have to rely on self-reported benchmarks from reps or look at metrics such as the number of deals an average rep closes to achieve these efficiency metrics.
A Growing Interest In Predictive AI
It turns out that while generative AI attracts much of the buzz around the technology, it is not the clear leader among CIOs in terms of driving value: 54% of tech leaders report they prioritize generative AI and 51% focus on automation, while 31% prioritize predictive AI. To capitalize on this discord and deliver value across a broad spectrum, AI models must be tuned to support workflow automation and predictive analytics.
“The different kinds of AI and how they are consumed also play a significant role,” said Reshef. “For instance, it’s easy to use a tool like ChatGPT to generate quick outputs in seconds. However, building a model capable of accurately predicting something as specific as quarterly business outcomes would require a proprietary solution, significant data configuration and extensive customization.”
Reshef continued that these two models have distinct trade-offs: Off-the-shelf models are accessible and efficient for general tasks, but transforming critical workflows demands deeper investment in data integration and domain-specific tuning.
Smaller Companies Are More Eager To Prove AI ROI
Smaller U.S. firms (250 to 500 employees) are more ROI-focused, with 40% willing to halt projects lacking clear ROI, compared to just 19% of larger companies. This suggests that while smaller U.S.— firms see the value in making investments in AI, they need to focus on initiatives that deliver measurable and immediate returns and have less budget for experimentation, whereas larger companies might have more capacity to invest in long-term projects without immediate ROI.
“Small businesses need to embrace an experimentation culture and set clear guardrails for those experiments’ timelines, financial expenditures and success metrics,” said Reshef. “More often than not, smaller companies must be more agile when assessing the return on their AI investment, while larger companies have the luxury of thinking more long-term. Prioritize the AI solutions that will improve a variety of business metrics rather than a single metric, and as your business grows, so will the flexibility of AI experimentation.”