What Happens Next

Lisa, Marcus, and David are not edge cases. They are the ordinary consequences of a system that governs through narrative. What changed is the cost of reconciliation. The first move is smaller than you think.

The structural problems described here are not rare pathologies. They are daily experience for millions of people inside software-dependent corporates: the people sitting in meetings they know are pointless, maintaining documents they know nobody trusts, and navigating structures that prevent them from doing the work they were hired to do.

Lisa leaves because telling the truth keeps getting classified as a communication problem. Marcus discovers that the knowledge he spent years using to hold a broken process together becomes more valuable once the process is finally owned. David still has a folder on his laptop labelled “synthesis” that he has never shared. Over time, the organisation will either make that private reconciliation operational, or it will continue to lose the people who carry it. The folder does not get smaller with time; the gap it documents gets wider, and carrying it becomes harder. Most software-dependent corporates have a David, and many of them are quietly updating their CVs.

Lisa, Marcus, and David are not edge cases. They are the ordinary human consequences of a system that governs through narrative instead of artefact.

What changed is not that executives suddenly became wiser or engineers suddenly became more strategic. What changed is that the cost of reconciliation collapsed. A machine can now compare what the organisation says against what the organisation runs often enough for the gap to matter.

That makes one question unavoidable: are you using AI to write a better story about the business, or to discover whether the story matches the business at all?

The first move is smaller than the transformation programmes of the last decade: one process, one owner, one boundary, one unit that can produce visible evidence fast enough to survive the organisation’s immune system. The Diagnostic Toolkit contains a board diagnostic, a 90-day starting sequence, and sample artefacts for the first unit.

The second move is political: find the executive who already knows the private reconciliation by heart and stop pretending it is private knowledge rather than organisational fact.

The competitor does not need to be smarter first. It needs to reduce the distance between what it says and what it actually runs.

What the first year looks like

The first year should not look like a company-wide transformation festival. No poster, slogan, launch video, or steering committee full of the very functions whose authority the change is going to reduce. That route neutralises the change before the change can produce evidence.

Pick one process that matters. Claims. Refunds. Order fulfilment. Customer onboarding. Something valuable enough that the results will matter and bounded enough that ownership can become real.

Name one owner. Give that owner actual authority, not ceremonial responsibility. Put the operational knowledge, the engineering capability, and the key decisions around that process inside one unit. Write down the process in a form precise enough to be tested against the system. Define the contracts to the adjacent units. Then protect the boundary long enough for the evidence to appear.

The board’s role in that first year is simple: hold the line long enough for reality to answer. Are meeting hours falling? Are changes shipping faster? Is ownership getting clearer rather than fuzzier? Are the people who care most about reality becoming more effective and more likely to stay?

If the answer is yes, the model earns the right to spread. If the answer is no, the model should be challenged on results, not smothered in advance by the same coordination machinery it was designed to replace.

You probably already know which side of the fork your organisation is on. The diagnosis is rarely the surprise. The surprise is discovering that the structural correction is smaller than the transformation programmes that preceded it, and that the first move is not a company-wide initiative but a single process, a single owner, and a willingness to let the evidence speak before the narrative has time to smooth it.

The gap between organisations that reconcile reality and organisations that narrate around it will become visible in cost, in speed, in truthfulness, and in the willingness of the best people to stay. Every reconciliation cycle one runs and the other does not gives the first structural knowledge the second will struggle to recover later. The brochureware companies had years to adjust. Organisations making the same mistake with AI are unlikely to get that much time.

← Where This Is Hard The Strategic Context Stack →