A central question in my MBA research with 18 UK SMEs was why some organisations made steady progress with GenAI while others stalled, despite similar constraints. The difference was not ambition, funding, or technical sophistication. It was leadership practice.
How higher-performing SMEs approached it
Higher-performing SMEs approached GenAI as an organisational learning challenge. Leaders framed GenAI in relation to purpose, not pressure. They were explicit about why it mattered and open about what was unknown. This created space for dialogue rather than speculation.
In these organisations, experimentation was collective rather than isolated. Leaders encouraged shared reflection and treated early use cases as opportunities for sensemaking, not proof of success. Governance conversations happened early — not to restrict activity, but to legitimise it.
How lower-performing SMEs approached it
Lower-performing SMEs exhibited a different pattern. Leaders focused heavily on tools and generic use cases, often delegating GenAI exploration to small technical groups or the IT department. Communication was limited, and uncertainty went unacknowledged.
In these contexts, GenAI activity appeared fragmented. Employees were unsure how experiments related to organisational priorities. Governance discussions were delayed until risks became visible, at which point they were experienced as restrictive rather than enabling.
The essential difference
The contrast was stark. High-success SMEs orchestrated meaning. Low-success SMEs optimised activity.
The implication is not that one group "cared more" about GenAI. It is that leaders who treated adoption as an interpretive process were better able to align action, trust, and learning. Progress did not depend on having the right answers. It depended on asking the right questions together.