What Companies with Successful AI Pilots Do Differently

According to a recent MIT report, a remarkable 95% of gen AI programs fail to deliver bottom-line returns. In the wake of that finding, most commentators focused their attention on trying to explain why so many programs fail. Nathan Furr and Andrew Shipilov, for example, recently highlighted for HBR the very real danger of the “experimentation trap,” in which pilots never connect to customer value or scale beyond the lab.

But after seeing MIT’s results, we got curious about a different question: What’s different about the 5% that succeed?

In search of an answer, we conducted dozens of executive interviews across Fortune 50 corporations, private-equity portfolio companies, and nonprofits. Most organizations, it turns out, have access to similar technologies and data, so that can’t explain the difference. Neither can access to better models or superior technical talent, because even modestly resourced firms sometimes outperform those with sought-after experts.

What we came to realize in the course of doing our interviews is that the differentiator is leadership.

When we asked 53 senior leaders to rank the single biggest driver of returns from AI, they confirmed this finding: 47% ranked leadership effectiveness as the single biggest driver of AI ROI, which put it far above all other options, including workflow integration (15%), organizational culture (11%), and engineering talent (8%). Fewer than half of those we interviewed, however, believed that their organizations had the right leaders in place to join that successful 5%.

In this article, drawing on what we’ve learned in our research and client engagements, we’ll discuss what makes companies successful in their AI efforts and will identify a set of specific leadership behaviors that separate the winners from the rest.

Read the article