Growth is often associated with expansion, diversification and increasing capability. But new research suggests the relationship is far less linear than most organizations assume.

A study led by researchers from MIT Sloan School of Management and the Santa Fe Institute found that as complex systems grow — from biological cells to corporations to cities — they tend to become more specialized rather than dramatically more diverse.

Published in the Proceedings of the National Academy of Sciences, the research identifies a common mathematical pattern governing how complexity evolves across very different systems. The core finding: growth usually leads to doing more of the same, not necessarily developing entirely new capabilities.

The researchers compared organizations to language systems using a concept known as Heaps’ Law — a principle showing that as texts become longer, the introduction of new words slows down. Applied to business, the same logic appears to hold: as companies scale, they increasingly hire for existing functions instead of creating new ones.

According to Vicky Chuqiao Yang, one of the study’s lead authors, this has important implications for how organizations approach transformation and innovation — particularly in areas such as AI adoption.

“If an organization wants to add a new function category like AI, you can’t simply hire one or two people and expect transformation to happen,” Yang explains in the research. “New capabilities require supporting infrastructure and the expansion of existing systems around them.”

The study suggests that complexity has thresholds. In practice, this means organizations often need to reach a certain scale, operational maturity and internal structure before entirely new functions can successfully emerge.

The findings also reinforce a broader leadership challenge increasingly visible across industries: scaling a company is not the same as evolving it.

While organizations naturally optimize for efficiency and repetition as they grow, meaningful innovation may require deliberate disruption of those patterns — alongside investments in culture, systems and cross-functional capability building.

In an era where companies are racing to integrate AI, redesign workflows and reinvent business models, the research offers a sobering reminder: becoming larger does not automatically mean becoming more adaptable.

—

Source: MIT Sloan