Where the Numbers Point
Ninety-five percent of enterprise AI initiatives fail to deliver, according to a figure reported by Fast Company this month. Behind it sits a UK-specific one: British businesses have already lost £67 billion on failed AI work, according to IT Brief UK.
The useful part of that story is the reason. Fast Company names it directly: the failures cluster around governance, not the models. The AI worked. The organisation around it was not yet ready to make use of it. That is good news, because readiness is something you can build.
The Models Are Not the Constraint
Frontier capability is now widely available and improving on a monthly cadence, open to any enterprise with a contract and a use case. If the model were the deciding factor, the results would track model quality. They track organisational readiness.
Readiness starts with something basic: being able to see the process an AI will work on. The procedure on paper and the process in practice drift apart in most organisations, and AI pointed at that gap will repeat whatever it finds there, quickly and at scale. Point it at a process that is written down, owned and current, and it can repeat what works.
That is why the 95% figure reads best as a governance statistic. The organisations in it did not lack capable models. They lacked a process foundation to place them on. The missing process layer is the upstream work that the successful minority did first.
What the 5% Do Differently
They reverse the sequence. They build process clarity before they automate anything.
That means writing down how the critical work is done and why: the steps, the decisions, the points where human judgement is needed, the controls that must hold. It is unglamorous work and it does not produce a demo. It produces a foundation that people and agents can both be held to.
The difference changes what AI is for you. On an unclear process, AI is a gamble, because you cannot predict where it will help. On a clear one, it becomes an investment you can plan: you know where it applies, you can measure whether it is working, and you can trace a problem to its cause. Our method is designed to produce exactly that: a living model of how the business is meant to run, not a diagram that ages on a shared drive.
Getting Your Next AI Spend Ready
If your board is pushing AI and your operations team is cautious, this sequence gives both of them what they need.
Step 1: Write down the critical processes AI will touch. Which processes matter most to the business? Start there. Define what they do, who owns each step, and what working correctly means.
Step 2: Make the decision logic explicit. Where are judgements made? What data drives them? Where must a person stay in control? These points become the human gates in any future automation.
Step 3: Choose where AI adds value first. Separate the places where automation creates measurable value from the places where it needs more preparation. This is a decision you can take with evidence.
Step 4: Deploy on a foundation that boards, regulators and auditors can follow. When the AI team receives the process, it receives clarity. The result is traceable and easy to defend.
IGX360 Insights shows you the provenance, confidence, freshness and ownership of the process knowledge you already hold, which is the starting point for this sequence. You can see the kind of output on a sample diagnostic before making any product decision.
Make the Next Pound Count
The £67 billion is what it costs to automate before the foundation is ready. It says nothing against AI as a bet. It says the sequence matters, and the sequence is within your control.
The board can back AI with confidence when it asks for one condition: process clarity before automation. That condition turns board caution into board confidence, and turns the next pound spent on AI into an investment.
You already know which processes your AI programme will touch. The next step is seeing them clearly enough to plan around them.
Talk to IGX about getting your first AI project ready to deliver.