Organisation & Governance

A simple diagnostic: when a boundary case appears, do people first consult the rules or first ask their manager? If it is the latter, the written rules are decoration and the real rule is a superior's momentary preference.

Three conditions make rules real: they are published and queryable; they cover high-frequency decisions rather than exotic ones; and their application leaves a trace, so the same case reaches the same conclusion whoever handles it.

The same question changes resource allocation depending on who it is asked about. For long-cycle industrial buyers the real priority order is usually supply reliability, then regulatory and audit support, then quality consistency, and only then price. Spending on the first three often locks a customer in more effectively than a price cut.

Internally, commission on first orders produces discount-driven volume with no repeat business; commission on margin protects price. Upstream, buying priority supply with committed volume and prompt payment in an annual framework beats relying on goodwill.

The value of the framing is that it converts a people problem into a parameter problem. The first is not actionable; the second is.

Regulatory filings and licences are entry barriers and belong in a central archive, with staff able to borrow but never hold originals. Client and decision-chain data belongs in one system, with account ownership in the company's name and access revoked the day someone leaves.

Expertise must be productised: turn one person's judgement into standard technical packages, FAQ manuals and comparison tables so less-experienced staff can apply most of it. Brand and content assets are the only compounding ones, because they generate enquiries without depending on any individual.

All four point the same way: organisational leverage is whatever survives turnover.

Execution work — documentation, order follow-up, shipping, collections, compliance review — suits strict procedure and zero tolerance for error, because its goal is the absence of surprises and it standardises well.

Creative work — new markets, new client types, technical solution design, content and brand — aims at surprises, and needs psychological safety plus a longer observation window. Manage it like execution and you get an organisation that only does what the rules already cover: a slow surrender of every new market.

So the two differ in review cadence, error tolerance, and decision authority. The one thing they share is that both must be reviewed afterwards.

This does not dissolve the organisation; it changes how the organisation is described. A traditional job description assumes a stable class of tasks requiring a fixed person for a fixed share of time. Once tasks can be decomposed, partly automated and recombined on demand, that description stops matching reality.

A more usable frame is matching task clusters against available capability — human judgement plus dispatchable automation — on a rolling basis. This also changes what turnover means: not betrayal, but a temporary gap in a capability network, where the crucial thing is retaining the knowledge assets that node carried.