Why AI projects struggle to move beyond the pilot stage

Most organisations no longer need convincing that AI has the potential to change how they operate. The pressure to explore AI capabilities is now coming from every direction, whether from boards, customers, employees, or wider market expectations. Across industries, businesses are investing time and budget into experimentation, often at pace.

What many organisations are now discovering, however, is that experimenting with AI and successfully embedding it into the business are two very different things.

A significant number of AI initiatives never progress beyond the pilot stage. In some cases, the technology itself works perfectly well, but momentum slows once questions around ownership, adoption, measurement, or operational value begin to emerge. Elsewhere, projects become disconnected from the original business challenge they were intended to address. Teams become fascinated by what the technology can do, without fully establishing whether it is solving the right problem in the first place.

These are not necessarily new challenges. Many previous waves of enterprise technology have faced similar issues. What feels different with AI is the speed at which organisations are expected to respond, often before they have had time to fully define what success looks like for them.

Moving from technology to outcome-focused conversations

At Version 1, this was one of the key reasons behind the development of our AI-enabled co-creation Studio in Dublin. The original intention was to create a collaborative innovation space for customers, but the concept evolved quite quickly into something more practical and outcome-focused.

Rather than building an environment centred on demonstrations or technology showcases, we wanted to create a space where organisations could step back and have more open conversations about the challenges they were trying to solve. The technology still plays an important role, but it is deliberately not the starting point.

When customers enter the studio, the discussion begins with their business context. We explore the friction points they are experiencing today, the pressures they are under, and what a better experience might look like for their employees, customers, or citizens. Sometimes organisations arrive with a very clear use case already in mind. Others simply know they need to begin somewhere but are unsure what the most valuable starting point might be.

That flexibility is important because every organisation is at a different stage of maturity with AI. Some need space for ideation and strategic thinking. Others need help connecting a defined business challenge with a realistic user experience or implementation path. In many cases, the role of the session is simply to create clarity and alignment across different stakeholders.

One of the most interesting pieces of feedback we receive is how valuable customers find the opportunity to focus on the business problem first. In technology environments, conversations can very quickly become dominated by architecture, tooling, or technical capability. While those considerations matter, they are often more productive once there is a shared understanding around the outcome the organisation wants to achieve.

AI encourages and supports collaboration

The studio has been purposefully designed to accelerate collaboration and exploration, as we did not want to risk removing the human element from the process. Throughout sessions, conversations are transcribed in real time, while AI agents work in the background to support the workshop activity. Depending on the needs of the session, they may generate research summaries, feasibility assessments, wireframes, architectural concepts, or early-stage demonstrators.

The intention is not to automate the workshop experience or replace human decision-making. The value comes from reducing the administrative overhead that often slows innovation projects down and allowing participants to focus more fully on discussion, challenge, and problem solving.

We have also introduced AI personas into the sessions, each designed to contribute a different perspective. Some represent product or research viewpoints, while others focus on areas such as responsible AI, sustainability, or critical challenge. In some workshops, personas are created to reflect the perspective of an end user or citizen group relevant to the discussion.

Interestingly, these contributions can help surface questions or concerns that participants may not naturally raise themselves. They also help broaden the conversation beyond purely technical considerations and encourage participants to think more carefully about adoption, trust, and long-term usability.

At the same time, facilitation remains central to the process. The technology is there to assist the conversation, not direct it. Experienced facilitators still play the key role in guiding discussions, interpreting challenges, and helping organisations navigate uncertainty.

Why adoption remains the hardest part

One of the recurring themes across AI projects is that implementation challenges are rarely confined to the technology itself. In many organisations, the more difficult work begins once a prototype has been created and the focus shifts toward operational adoption.

Questions around user confidence, behavioural change, governance, and measurement often become far more significant at that stage. Employees may feel uncertain about how AI will affect their roles. Business leaders may struggle to define meaningful success metrics. Different departments may have conflicting priorities or expectations about what the technology should deliver. These factors can slow progress considerably if they are not addressed early.

For that reason, we place significant emphasis on bringing the right mix of stakeholders into the co-creation process from the outset. That includes business users, technical specialists, operational teams, and leadership representatives. The aim is to create shared ownership around both the problem and the proposed path forward.

Equally important is establishing how success will be measured before substantial investment is made. Organisations do not always need fully formed answers at the beginning, but they do need a clear understanding of the value they are trying to create and why it matters strategically.

Without that clarity, projects can easily lose direction over time. It becomes difficult to maintain momentum, secure further investment, or build confidence internally around scaling adoption. 

Creating a clear path to implementation

The studio model was never intended to suggest that complex AI transformation can happen in a single workshop. In practice, successful delivery still requires discovery, user engagement, governance, implementation planning, and operational change management.

What the process can do, however, is shorten the distance between an initial conversation and a tangible direction of travel. Organisations leave with something concrete, whether that is a demonstrator, a user concept, a strategic framework, or a clearer understanding of the challenge itself. That creates momentum and gives teams something practical around which they can align future decision-making.

In some cases, customers have subsequently asked about building similar environments within their own organisations, something that was never part of the original plan. What that reflects, perhaps, is a wider recognition that businesses need more effective ways to bring together technology, strategy, and people in one place.

As organisations continue exploring how AI fits into their long-term operations, there is likely to be increasing emphasis on approaches that prioritise collaboration, clarity, and measurable value alongside technical innovation. The organisations making the strongest progress are often those creating space for experimentation while remaining closely connected to real operational and human challenges.