THE PRACTICES
Practice I:
Corporate Strategy & M&A Decision Architecture
Practice II:
Applied Behavioral Design & Choice Architecture
Practice III:
Bionic Operating Models & AI Transformation
Practice IV:
Executive Sounding Board & High-Stakes Leadership
What breaks.
Deal conviction forms early and hardens fast. By the time diligence reports, the committee is defending a position rather than testing one. Escalation of commitment is the most expensive bias in corporate finance and the least examined, because it looks identical to leadership.
How we fix it.
Pre-mortems run before the banker's pitch, not after. Structured dissent with named roles and real air cover. Reference-class forecasting against outcomes rather than plans. Decision logs that separate the quality of the choice from the quality of the result, so the committee can learn from a good decision that lost money and distrust a bad one that made it.
What it is worth.
Fewer deals, higher conviction, faster to no. The value is disproportionately in the transactions you decline in week three instead of month five, and in the diligence cost you never incur.
What breaks.
The strategy assumed people would act on new information. They almost never do. Friction, defaults and social proof determine behaviour far more reliably than intent, and almost no operating model is designed around that fact.
How we fix it.
Map the friction.
Redesign the defaults. Build the experiment.
Run it small, measure honestly, kill what does not work, scale what does. Behavioural design is only credible when it is falsifiable, which is precisely why so little of it survives contact with a P&L.
What it is worth.
Conversion, adoption and retention moved by measurable margins on tested interventions, typically at a fraction of the cost of the campaign or system rebuild it replaces. Every claim carries a control group.
What breaks.
The technology is not the constraint and has not been for some time. The constraint is accountability. When a model makes a recommendation, someone must own the outcome, and most organisations have quietly left that seat empty.
How we fix it.
Decision rights redrawn first, technology second. Identify which decisions should be automated, which augmented and which must stay stubbornly human. Rebuild the escalation paths. Design the trust calibration, because a workforce that over-trusts a model is as dangerous as one that ignores it.
What it is worth.
Cycle time on recurring decisions, measured before and after. Reduced rework. Most valuably, an AI investment that appears in the operating result rather than only in the capital plan.
What breaks.
Seniority narrows counsel. The chief executive's information is filtered by people managing their own exposure, and the board sees a version prepared for the board. The most consequential decisions are therefore taken with the least honest inputs in the organisation.
How we fix it.
A standing, confidential counterparty with the operating scars to be useful and no internal position to defend.
Board paper stress-testing. Rehearsal and sparring for the conversations and decisions that matter. Occasionally, the observation nobody on the payroll can afford or is willing to make.
What it is worth. Unpriceable, which is why it is retained long term relationship rather than only billed by the day. Boards that are candid with themselves outperform boards that are polite, often with an independent instigator can help the conversation and decisions.