A written win condition: how a first engagement is scoped and scored

Most software engagements are judged in hindsight, by feel, at renewal time. Ours are judged against a document written before anything runs. This note explains why we insist on that, what the document contains, and how it is scored, so you know what you are agreeing to when you take a demo call.

Why write it down first

A decision intelligence system produces forecasts, risk assessments, and scenarios. All three are probabilistic, which means any single outcome can be argued either way after the fact. The only way to know whether the system helped is to agree in advance on what "helped" means and to measure it in the open.

This is the same discipline the system applies to itself. Forecasts are registered before the outcome and scored against baselines with no hindsight edits. An engagement should be held to no lower a standard than the software.

What goes in the win condition

The document is short, usually a page, and it is written together on and after the demo call. It names four things.

• The decisions in scope. One or two, described as the people who make them describe them: which supplier to lean on this quarter, whether to pre-position inventory before the season, how to respond to a lane disruption.
• The baseline. How those decisions are made today, and what a "no change" outcome looks like, so improvement is measured against something real rather than against zero.
• The measures. What will be counted, over what period, and how. Time from event to a decision-maker being informed. Forecasts scored against the baseline. Scenarios run before a decision rather than after. Whatever fits the decisions, stated concretely.
• The stop conditions. What would make either side end the engagement early, written down while everyone is calm.

How it is scored

Scoring is done in the open, on a schedule agreed in the document, from the system's own records. Registered forecasts and their scores, the events that reached the decision-makers and when, the scenarios that were run and what was decided. Both sides read the same numbers. Where the system did not help, the score says so, and that is treated as information about where to point it next rather than as something to explain away.

What this changes for you

You know before you start what you are paying for, how it will be judged, and when you can stop. Pricing follows the same document: engagements are priced against the decisions in scope, the people who will use the system, and the connectors the deployment needs, with no public rate card and no automated checkout, because the scope is what determines the price.

Where SABLE stands

A first engagement starts on the decisions where the outside world already matters, since the external intelligence stack, the knowledge graph, and the forecasting, risk, and scenario engines are live today, while governed connectors into a company's own systems are in development. Forecasts and scenarios are probabilistic model outputs provided for informational purposes only, and the decision stays with the person who owns it. The written win condition is how both sides keep that honest.