Agile Analysis
Agile Analysis, as defined in the Agile Extension to the BABOK Guide, is the subject of this section. The IIBA-AAC exam tests it through scenario questions that ask whether you treat analysis as continuous, collaborative and right-sized across three overlapping planning horizons rather than performed once at project start.
Three horizons, not three phases
Strategic answers why change is needed: business need, current and future state, a prioritised backlog of ideas, and value measures such as KPIs. Initiative answers what to do: initiative vision, roadmap, stakeholder map, product backlog, hypotheses, MVP scope. Delivery answers how the team builds it: stories, acceptance criteria, example mapping, Definition of Done, prioritisation. The horizons run concurrently; a delivery insight can send you back to strategy.
Analyse just in time, just enough
Reduce uncertainty and defer irreversible decisions to the last responsible moment. Prefer real examples, models and conversations over speculative documentation. Apply the seven guiding principles: see the whole; think as a customer; analyse to determine what is valuable; get real using examples; understand what is doable; stimulate collaboration and continuous improvement; avoid waste. Use the six BACCM concepts - change, need, solution, stakeholder, value, context - to check a proposed change is coherent.
Worked example
A retailer's sponsor asks for a loyalty app by Q4. Strategic analysis uncovers the real need: repeat-purchase rate is falling. Initiative analysis frames two options, states a testable hypothesis ("targeted offers lift repeat purchase 10%") and slices an MVP. Delivery analysis writes acceptance criteria and runs example mapping with the sponsor. Six weeks later the data refutes the hypothesis, so the team returns to the initiative horizon. The request was a solution; analysis moved up a horizon to find the need, then down to make delivery testable.
Traps that cost marks
Treating horizons as sequential phases; producing an exhaustive requirements document "just in case"; reporting outputs (features shipped) instead of outcomes; skipping strategy because a solution was already named; and missing "see the whole" by optimising one team while the value proposition fails.