The standard objection to any whole-business intelligence proposal is reasonable: we do not have the budget or appetite for an enterprise programme.
Correct. And you do not need one.
The false choice
Organizations are usually offered two options.
Narrow and shallow. Solve one team's reporting problem. Fast, cheap, and it does not touch the cross-functional questions — which are the ones that were hard.
Wide and heavy. An enterprise data programme. Eighteen months, a platform decision, a governance function, and value arriving after most of the budget.
Both are real. Neither is what most firms need, because the actual requirement is narrow in scope and wide in context.
What the third option looks like
Take one management question. Map the enterprise around it.
One question — why is delivery margin falling — but mapped across every function that bears on it: finance, workforce, delivery, sales. Four domains, not twenty. The systems that matter for that question, not every system in the estate.
What you get is a genuine cross-functional model, at a scope that fits in weeks rather than quarters, and an answer to a question that was actually being asked.
Why this works better than it sounds
Three reasons.
The entities repeat. The definitions you must settle to answer a margin question — project, person, customer, period — are the same ones needed for most other cross-functional questions. Settling them once for one question resolves a large share of the next one.
The question forces prioritization. An enterprise programme must decide what matters in the abstract, which is close to impossible and produces scope that expands. A specific question decides for you: if it does not bear on the question, it is out.
It produces evidence rather than a plan. At the end you can answer something, with real data, that could not be answered before. That is a much stronger basis for the next investment decision than a roadmap.
What it is not
It is not a pilot in the sense of a disposable trial. The model, the definitions, and the connections built for the first question are the foundation for the second — provided they are built as reusable assets rather than as a one-off analysis.
That distinction is the whole difference between a proof that compounds and a proof that gets thrown away.
The test
If a proposal requires you to centralize the data estate before it can answer anything, it is the wide-and-heavy option regardless of how it is framed.
If it answers one real question and leaves behind entities and definitions the next question can reuse, it is the third option.
Practical check
The reuse test is simple: after the first question is answered, identify which entities, definitions, and join paths the next question can inherit. If nothing carries forward, the work was narrow and disposable. If the foundation reduces the cost of the next answer, the scope was narrow and the context was genuinely wide.
The Zenkei Method is built on this principle, and a 360° Enterprise Intelligence Blueprint is its first application.