There's no doubt that AI is the future of business. It's being embedded in every industry you can think of, from hospitality and finance to healthcare and education. With use cases proliferating and tools growing more powerful with each passing day, the pressure to "do something" with AI is now almost universal.
But despite its growing ubiquity, many businesses are struggling to see any meaningful returns on their AI investments, especially where monetisation is concerned.
This gap between investment and outcomes is driven, in part, by businesses' laser focus on the technology and neglect of the people required to operate it. When it comes to a successful AI strategy, people are critical.
Therefore, gaining clear oversight of people-focused AI metrics must be at the top of the agenda for boards as they look to improve AI outcomes and deliver ROI. This must include oversight of AI's adoption throughout the workforce, and training.
But before boards can focus on training and adoption, they must gain clear oversight over AI outcomes. Just as you'd judge a plane on its ability to transport passengers safely to a new location and not on the curvature of its wings, AI must be assessed on its ability to deliver results, and not on its inherent or anecdotal technical qualities.
The impact of AI strategy and initiatives on metrics like productivity gains, cost savings, and improved customer experience must be closely monitored and linked directly to profit and loss statements – where the success or failure of AI must ultimately be judged. Boards must also be active in setting desired KPI outcomes for the technology and measure progress against these to ensure clear expectations are in place.
Without oversight of these metrics, boards won't be able to identify where and why the AI strategy isn't having its desired results, and upgrade the people-focused side of the strategy accordingly.
The first people-focused metric boards must gain oversight of is adoption – the rate and extent at which employees are adopting the AI tools and procedures they're equipped with.
Tracking usage patterns provides a useful baseline, enabling boards to pinpoint where uptake is missing, patchy, or infrequent, and follow up with initiatives for increasing uptake.
This should extend beyond simply tallying the number of employees using AI but rather take an in-depth look at AI usage across teams, departments, and all seniorities to gain a more comprehensive understanding of where, and by whom, AI is being used. Usage at the senior, strategic level is perhaps most critical, and shouldn't be forgotten.
Boards must also monitor how frequently AI is being used by employees, because it's one thing for employees to report that they're using AI – but if usage is infrequent or ad hoc, the value will be minimal.
The second metric is AI training – a composite of how effectively the workforce is being upskilled and reskilled in AI technologies and the skills they require.
In short, it's no good pushing for widespread AI adoption if no one knows how to utilise the tools effectively. However, it's equally of little use for businesses to assign employees a single training task and then leave them to their own devices. Training must provide employees with a clear understanding of how AI can be applied in their role, and it must leave them feeling confident and capable of effectively deploying AI tools.
Low confidence in using AI is just as much of a barrier to widespread adoption as a lack of training. That's why, beyond tracking who has been trained and when, boards should also measure how confident staff feel about using AI in their roles. Regular staff surveys or check-ins can help build a clearer picture of employee attitudes, and when combined with adoption data, identify areas where additional support is needed.
The workforce needs to be trained, supported, and empowered to utilise AI in their daily work. Tracking training, and then monitoring adoption and usage patterns, will give boards the insight they need to evaluate the strength of the people in their AI strategy – and their impact on AI outcomes.
AI transformation is a marathon, not a sprint – but businesses risk tripping themselves up by treating it as a shortcut to instant value creation. Simply buying a licence for the latest AI platform won't deliver value on its own. The real impact comes when investments in tools are matched by investments in people.
