Builds the alternative set a buyer would weigh from research alone.
Reads the market, competitive and customer-profile research and builds the set of alternatives a buyer in this market would weigh, clustered into three to five groups named in the research's own language, with do-nothing, the manual workaround and the in-house build carried as clusters wherever the research supports them. Every row is tagged research-derived, so what it produces is the supply the research lists — not a record of what buyers actually chose.
Its purpose is to give the attribute and segment work an alternative set that states its own limits. The blind spot is written into the record rather than a footnote — no frequency or share is ever claimed, and do-nothing is systematically under-represented — and any competitor on the internal list that the research does not support is recorded as positioning baggage rather than quietly kept.
What it needs, and which agents use what it produces.