Real ROI from AI: strategy plus capability

The companies seeing real ROI from AI do two things well:

1. They are clear about what AI is for.
2. And they give their people the ability to use it.

That combination shows up repeatedly in the 2026 research:

• The biggest barrier to integrating AI into real workflows is not technology, but workforce skills (Deloitte's State of AI 2026)
• Employees whose managers actively support AI use are far more likely to use it frequently, and far more likely to say it improves the quality of their work. (Gallup 2026 research)

Core takeaways:

• Clear strategy tells people where AI should help, and where it should not.
• Training gives people the confidence to use it well in their daily work.

One without the other does not compound.

If you want AI to do real work, start by being clear about the work you want it to do. Then make sure your people are able to work with it, and be intentional about what you expect from your teams (for example, measure time saved on routine tasks).

In a practical way this means identifying work your teams currently perform, and using GenAI to augment their work (help them do the work at the same quality, but more efficiently). Then you use the saved hours to focus your humans on work that AI does poorly: time with customers, strategy planning, making real connections.