A friend sent me a William Blair research report this week that characterized enterprise AI adoption as “a mile wide and an inch deep.”
Just about every organization has started to use AI tools in some fashion. But there’s little orchestration of how those tools are actually used. And very little alignment across the enterprise.
A 2025 MIT report suggested that 95% of enterprise Generative AI pilots failed to deliver significant ROI or move beyond experimentation. The report flags not the quality of the tools, but to strategic misalignment as the primary driver.
Of course, strategic misalignment is an age-old challenge for organizations. AI acceleration just further exposes the rift.
Hugh Derrick at eatbigfish recently pointed to Harvard Business Review research that strategic alignment is up to 3X lower than leaders think. As he put it:
“So when you layer AI-driven speed on top, you don’t magically become more effective. You run the risk of getting faster at being inconsistent (and yes, you’ll create a lot more ‘stuff’ along the way)…
“And it matters most in big, complex organizations where silos already slow everything down. When 30% of senior leaders point to silos as a root cause of productivity stagnation, and two‑thirds say their organizations are overly complex, adding more speed to a fragmented system doesn’t fix the system. It just creates more noise.”
Here are a few related cartoons I’ve drawn over the years:
