The conversation is no longer about whether AI matters. That question has largely been answered. For most businesses I speak to, AI is already part of everyday working life. It is drafting content, analysing data, supporting customer queries, speeding up admin and helping teams think more creatively.
What is becoming clear, however, is that while activity is high, confidence is often not.
Through our work with SME leaders over the past year, and the AI readiness research we carried out towards the end of 2025, one pattern has come through very clearly. Many organisations are actively experimenting with AI, often across multiple teams, but without a clear, shared sense of direction. AI is in use, but it is not yet anchored to strategy, priorities or measurable outcomes.
This is not a failure. It is a natural stage of progress. But it does place many leadership teams at an important crossroads.
When experimentation starts to plateau
In the early stages of AI adoption, experimentation is exactly the right place to start. It allows people to explore what is possible, build confidence and test ideas without heavy investment or risk.
That is what many SMEs have done well.
What emerged clearly from the survey data, however, is that experimentation often happens in pockets. Marketing teams adopt one set of tools, operations another, and individuals bring in their own solutions. Very quickly, the picture becomes fragmented. Leaders told us they could see AI being used but struggled to articulate where value was being created across the business as a whole, or how those individual efforts connected.
This is where activity can feel busy rather than impactful. Teams are doing more, but leaders are unsure whether the business is genuinely moving forward. Several respondents described this stage as frustrating rather than exciting. They could see the potential of AI, but they were unsure how it linked to wider business priorities and ROI.
The hidden cost of staying in “trial mode”
Remaining in trial mode can feel like the safest option, particularly given the pace at which AI tools continue to evolve. But staying here for too long carries real costs. When AI use remains informal and uncoordinated, value is difficult to scale. Time savings achieved in one area are rarely replicated elsewhere, learning is duplicated and processes are improved in isolation rather than end to end. Leaders told us they could see value emerging but lacked the structure to capture it consistently.
There is also a growing people and investment risk. Many leaders acknowledged that staff were already using AI, sometimes without clear guidance or oversight. They were unsure how widespread AI use was or whether expectations around data, quality and accountability were understood. At the same time, uncertainty around cost versus value continues to stall progress. Some organisations are planning sizeable investment, others have nothing formally planned, but in both cases the issue is rarely resistance to AI itself. It is a lack of confidence that investment will be purposeful and deliver measurable returns.
What execution actually looks like for SMEs
When I talk about execution, I am not referring to large-scale transformation programmes or expensive, all-in-one technology rollouts. For most SMEs, effective AI execution is far more focused and proportionate.
What comes through clearly from the research and from conversations with leaders is that progress tends to come from a small number of deliberate shifts:
From tools to outcomes: Leaders who are moving forward start with business challenges rather than technology. They focus on where time is being lost, where decisions could be better informed, or where capacity could be freed up, and then assess whether AI can play a meaningful role.
From individual use to shared direction: This does not mean shutting down experimentation, but it does mean providing clarity. In his article for Harvard Business Review, David De Cremer said that, “The savviest leaders prioritize participation by the rank and file throughout the adoption process.” In practice this can look like guardrails, shared priorities and agreed ways of working allowing teams to innovate without creating risk or inconsistency.
From open-ended experimentation to defined pilots: Rather than trying everything, organisations focus on a small number of low-risk, high-impact use cases. These are scoped carefully, supported properly and measured consistently, building evidence, learning and confidence.
At this point, it becomes clear that execution is not primarily a technology challenge. It is a leadership one.
The leadership role at this point in the journey
One of the strongest messages from the research was that the biggest challenges around AI are rarely technical.
They are strategic.
Leaders do not need to become AI experts. But they do need to take ownership of direction. That means moving AI out of ad hoc experimentation and into the wider leadership conversation. It means being clear about how AI supports business strategy, how it aligns with organisational values and how risks are managed responsibly.
Many survey respondents expressed a clear appetite for structured support at this point. Not more tools, but help with prioritisation, decision-making and alignment. They wanted space to pause, step back and agree a sensible path forward before scaling further.
Without that pause, organisations risk either investing without clarity or hesitating indefinitely.
Three questions I encourage leadership teams to answer now
Before investing further in AI, I encourage leadership teams to spend time together answering three questions:
Where could AI create the most meaningful value for our business over the next 12 to 18 months?
This needs to be grounded in your actual processes, customers and challenges, not abstract ambition.
What guardrails do we need to ensure AI is used safely, ethically and consistently?
Including expectations around data, quality and accountability.
How will we know whether AI adoption is genuinely delivering value?
What does success look like in practical, business terms?
These are not technology questions. They are leadership questions.
From curiosity to confidence
What the research reinforced for me is that SME leaders are not hesitant about AI. They are careful. They want to make good decisions, not fast ones, and they want to avoid wasted effort as much as missed opportunity. Does this resonate with you?
The organisations that will gain the most from AI over the coming year will not be those chasing every new tool, but those grounding their approach in strategy, leadership alignment and practical outcomes.
Moving from experimentation to execution is not about accelerating adoption. It is about slowing down just enough to choose the right direction.
For many leaders, and business teams, that moment of readiness is the most important step of all. At Alcea, this is where we work best, supporting leaders and teams to build clarity, define priorities and lead AI adoption with beneficial outcomes and confidence.
