The AI Value Gap: Why Businesses Need to Get Better at Stopping

Businesses are pouring money into AI and new transformation programmes, often treating the number of initiatives launched as a sign of progress. But launching more work does not guarantee better results.

The real challenge is knowing which initiatives are creating value and which need to change or stop. This requires a clear view of the portfolio, with decisions based on evidence and people with the skills to deliver. Without these foundations, the gap between ambition and execution will continue to grow.

Understanding the Value Gap

According to Emergn’s Value Gap survey report of 700 senior leaders in the UK and US, organisations lose an average of 2.4% of annual revenue to transformations that fail to deliver. In addition, 42% of leaders are managing between seven and ten initiatives at the same time. This might look like ambition, but in reality, it can be a sign of an organisation trying to do too much at once, spreading its resources too thinly and ultimately paying for it in lost revenue.

Leaders are not short of ambition. Most already believe they have clear processes and accountability in place. Yet transformations still fail to deliver as expected. Money, time and employee effort continue to be invested, even when there is limited evidence that an initiative will create value. The problem is not simply starting the right work, but recognising when the evidence says it is time to change course or stop.

Why Failing Initiatives Continue

Seven in ten organisations do not stop underperforming AI initiatives at the right time, according to Emergn’s research. They either act only after significant losses have already occurred or do not stop at all. Only three in ten leaders say stopping underperforming work is a normal part of how their organisation operates.

That is the bigger problem. Most organisations have not built stopping into the way they work. Instead, they treat it as an emergency measure rather than a routine business decision. It should be the other way around.

For 41% of leaders, an initiative is only stopped after significant time and money have already been lost. Almost a quarter also report that projects continue because too much has already been invested. In these situations, sunk costs and internal politics can become more influential than evidence.

This is not simply a data problem. It is a courage problem. Leaders need to be willing to challenge decisions they have already made when the evidence changes, rather than continuing to defend the original investment.

Stopping an initiative should not be seen as a failure. Done at the right time, it protects resources and allows teams to focus on work that can deliver measurable value. But leaders can only make these decisions early if they have a clear view of what is happening across the organisation.

The Role of Governance

Emergn’s research found that 71% of leaders cannot immediately provide the board with a complete, up-to-date view of every transformation and AI initiative. The information can take days or weeks to assemble, while some organisations would struggle to produce an accurate view at all. I would go further: if a board cannot get a straight answer about what is actually running right now, governance is not simply weak. It is absent.

The issue is even more pronounced in the UK. Only 7% of UK leaders say every initiative is formally tracked and reported, compared with 20% in the US. That gap should concern UK boards. It suggests that many UK organisations are less equipped to identify underperforming work early, challenge investment decisions and act before significant value is lost.

Poor visibility makes it harder to identify problems and compare priorities. It also spreads teams and budgets across too many initiatives, allowing failing work to continue while stronger ideas struggle to gain momentum.

Reporting culture can make the problem worse. Roughly one in five leaders say reports present a more positive picture than reality, bad news is softened, or people remain silent about programmes they believe are failing. Leaders cannot govern work they cannot see, particularly when the information reaching them is incomplete or overly optimistic. In my experience, a portfolio that always looks fine is often the one leaders should be questioning most.

Closing the Gap

Organisations need to connect funding to measurable outcomes. Every initiative should have a clear purpose, with its assumptions tested throughout delivery. Continued investment should depend on evidence that the work is creating value, not simply on the fact that time and money have already been committed.

That requires a live view of the portfolio. Leaders need current information on what is running, how it is progressing and whether it is delivering. Clear review points can then help them decide whether work should continue, change or stop.

Organisations also need to build capability within their teams. Learning should be part of everyday work, rather than relying on one-off training or a small number of experienced people. This is particularly important for AI, where successful delivery depends on judgement, problem framing and a relentless focus on outcomes.

Moving Forward

Four in five leaders already recognise the importance of product management and a product-centric mindset. The challenge is putting those ideas into practice. Recognising the principle costs nothing. Building the discipline to act on it is where many organisations stall and where the real competitive advantage now lies.

Starting another AI initiative is easy. Stopping one is harder. But the number of initiatives launched does not measure progress. What matters is whether organisations can see where value is being created, recognise when it is not, and have the discipline to act when the evidence changes.

The organisations that get this right will not necessarily be the ones doing the most. They will be the ones making better decisions about what to start, what to continue and, crucially, what to stop.

‍

‍