Most managers, when confronted with a difficult strategic decision, would like to know what is the prevailing practice in their industry. Prevailing practice interests managers, in part, because it may indicate which decisions are good, even “best.” The rationale is “Darwinian economics,” which proposes that competitive markets operate to force at least an approximation of optimal behavior, because “wrong” decisions are extinguished by the market mechanism. If so, the bulk of survivors make good business decisions and their collective practice is worth knowing, meaning it should be difficult to improve upon. This paper tests whether prevailing practice represents “good” decisions. The context is the decision whether to fill a sales district with a company (integrated) sales force or with outside agents (independents). The outcome of the choice is the ratio of sales to the cost of running the sales force, i.e., selling efficiency. A model is described which summarizes industry decisions, and two propositions are developed. Proposition 1, a simple version, states that efficiency declines as firms depart from the industry rule as to when they should integrate. Proposition 2, a contingent version, states that efficiency declines as firms depart from the industry rule in unpredictable selling environments but that conforming to the rule has no impact in predictable environments. The rationale is that firms learn the environment well enough to know when to depart from the rule in predictable settings but not otherwise. The relationship between efficiency and conformity to prevailing practice is examined for 93 sales districts of 11 recognized electronic component manufacturers. The results support proposition 2: the most efficient sales forces conform to prevailing practice and the least efficient deviate, but only in uncertain environments. In more certain environments, conformity and efficiency are unrelated. Implications (such as when it pays to be different) and suggestions for future research are discussed.
Stored only when both papers are in the corpus — 99 citations exist upstream.
Imported from OpenAlex (CC0) — nothing here is generated.