The Statistic Everyone Is Quoting
This week CGI warned that AI adoption is outrunning organisational readiness. A separate report put a number on it: only 1 in 10 organisations believe they are structured to capture AI value.
Read it the other way and it is an invitation. The ten percent who are ready have made a series of deliberate choices, and those choices are open to everyone.
The boardroom question has moved on. Twelve months ago it was whether to invest in AI. Now it is how to make the investment pay off at scale. The pilots looked promising and the budget was approved. The next step is the structure that lets the value arrive.
Readiness Means a Process AI Can Act On
When analysts say an organisation is not yet ready for AI, they are not commenting on its access to frontier models, which are available to everyone. They are saying the operating model needs to be explicit enough for AI to work within.
Conway’s Law states that the systems an organisation builds mirror the way it is structured to communicate. The corollary, as one recent analysis put it, is that your operating model matters more than your AI model. An AI agent works well on processes that are written down, current and owned. It works less well on tribal knowledge, scattered spreadsheets and procedures three years out of date, because it inherits whatever structure it is given.
This is not a single voice. Argon & Co and Lancia Consult, working independently, arrive at the same conclusion: operating-model transformation and discovery-phase work are prerequisites to unlocking AI and ERP value. A consensus is forming, and it points at the structure around the model. That is the basis of our method, and it is why we argue that process intelligence is the prerequisite for AI.
Why Pilots Stall, and How to Move Them On
A pilot succeeds in controlled conditions. The process is narrow, the data is clean, the edge cases are excluded and the team knows the workflow by heart. Scaling meets the wider process estate, with its exceptions that live in people’s heads.
To scale an AI initiative across your estate, the agent needs something to generalise from. Each process you write down gives every later deployment a head start, so the effort compounds in your favour. The alternative is a fresh discovery exercise for each deployment, and the cost of that compounds too. The Platform is built to make the first route the easy one.
Write Down How the Work Runs
Process articulation is the explicit, decision-level definition of how work is done: what steps occur, what decisions get made, who holds authority at each point and what data flows where. It is a clear model of the operating reality, in place of a narrative procedure or a diagram that ages on a shared drive.
An articulated process gives an agent something coherent to reason about, a place where human authority is explicitly required and a baseline against which value can be measured.
It also turns readiness from a vague aspiration into a measurable target. You stop asking whether you are ready for AI and start asking which processes are written down, which are not, and what it would take to close the difference. That question has an answer. The Platform is where the articulation happens, and our method is how we get there.
Be in the 1 in 10 by Your Next Board Review
Your peers are reading the same statistics, and your board may already have seen the 1-in-10 figure. Here is the encouraging part. Being in the 1 in 10 is a process decision, and the enterprises that capture AI value are the ones whose operating model is clear enough to act on. You can build that deliberately, starting with the processes that matter most. It is the work we do for transformation leaders.
A useful question for your next review: how much of your process estate could AI act on today, and can you show the evidence?