Foundations, Terminology and Scope (provisional)
Foundations
Why the foundations matter
Every organised effort turns inputs into outcomes for someone. The foundational layer supplies the vocabulary, lifecycle and value logic that everything else depends on: what problem is being solved, who benefits, what "done" looks like, and who is accountable for each decision.
Concepts to hold precisely
- Output, outcome, benefit. An output is a deliverable (a report, a system, a procedure). An outcome is a change in capability or behaviour. A benefit is measurable value realised from that change. Questions routinely swap one for another.
- The lifecycle. Concept, analysis, design, build, transition, operate, retire. Methodologies rename and reorder these phases, but items in a question usually belong cleanly to one of them.
- Roles and accountability. The sponsor funds the work and owns the benefit case; the business owner owns the process or product after handover; the analyst elicits and structures; the delivery team builds; assurance provides an independent check.
- The business case. Options, costs, benefits, risks and the reasoning linking them. It is a decision aid, not a funding formality.
Traps that recur
Confusing project success (on time, on budget, to specification) with benefit realisation, which is only observable after handover. Assuming the sponsor is the person who uses the result. Treating "requirements" as a synonym for "features". Assuming a phase must be fully complete before the next begins when the methodology in the scenario is explicitly iterative.
Rule of thumb: the "so what?" test
If you cannot name who receives the output, what changes for them, and how you would observe that change within a stated period, the foundation of the work is unfinished. Apply this to any proposed deliverable before accepting it as defined.
Worked illustration
A team reports "training delivered to 200 staff" (output), "staff now use the new process without support" (outcome), and "processing time down 15%" (benefit). A question asking which statement proves value has been realised accepts only the third.
Terminology and Scope
Precision is the subject
This domain tests whether you can pin down meaning. A shared glossary prevents two teams using one word for two things, and a written scope boundary prevents unmanaged expansion.
Concepts to hold
- In scope, out of scope, and deferred. Out-of-scope statements are active content, not omissions; deferred items go to a named backlog with a decision owner.
- Context diagram. Shows the thing being changed, its external interfaces and the flows across each boundary. It is the cheapest defence against missed stakeholders.
- Assumption, constraint, dependency. Assumption: believed true, not yet verified. Constraint: a fixed limit. Dependency: an external condition you need someone else to satisfy. Questions shuffle these three constantly.
- Normative words. "Shall" is mandatory, "should" is recommended, "may" is optional. Ambiguous modal verbs are the most common defect in requirements.
- Scope creep versus gold plating. Creep is uncontrolled addition from outside; gold plating is unrequested embellishment by the delivery team. Both damage value, but the source differs.
Traps
Accepting a scope statement made of nouns ("the reporting module") without behaviour. Confusing a deferred item with an out-of-scope item. Assuming a verbal agreement changes scope without a recorded decision.
Rule of thumb: the boundary test
Every scope statement should be readable as "for [who], the work delivers [what] which enables [outcome], explicitly excluding [what]". If the exclusion clause is empty, treat the scope as at risk.
Worked illustration
"The system should be fast." Rewrite as: "For warehouse supervisors, the system shall return stock queries in under two seconds for 95% of requests, excluding bulk exports." Note the actor, the mandatory verb, the measurable threshold and the explicit exclusion.