From a research thesis
to a decision system.

The first code was built inside a study of customer retention. It gathered reviews, observations, sales records and competitor evidence, then organised them into four domains: service, atmosphere, food and drinks, and value for money.

The workflow compared Pahat Pahit with its peers, estimated relationships between the domains, simulated an improvement in each one and selected the intervention with the greatest apparent opportunity.

Its world was deliberately small: one organisation, four levers and one pressing management question. But it established the idea that would endure.

The second form began when intervention ranking proved too narrow. Improving a customer-facing score could not explain whether an organisation possessed the capacity to convert the resulting demand.

The code's world expanded. Strategic resources were divided between outward activation and the structures required to sustain it. Conversion capacity, visibility, referral propagation and market power became explicit conditions. Goodwill could accumulate, persist and decay.

Most importantly, activation acquired a ceiling. Attention beyond the organisation's absorptive capacity could become congestion rather than growth.

The answer was no longer a single lever. It became a balance that changed with the state of the organisation.

The third form emerged when the model itself was no longer enough. A defensible decision required its surrounding evidence, assumptions, alternatives, uncertainties and history to remain connected.

The code reorganised itself around an Investigation. Evidence became a traceable artefact. Alternatives became explicit. Modelling, simulation, robustness and decision acquired separate readiness states. Each capability received only the context it needed. Success and failure entered a shared history, while unsupported conclusions were prevented from appearing complete.

Competitors changed from rows in a comparison table into actors whose behaviour could be simulated, observed and recalibrated. The final Brief stopped being merely the last output of a pipeline. It became the earned synthesis of what the investigation could support.