C onventional wisdom, largely based on PIMS data, nudges a brand manager toward the pursuit of market share even if it comes at the cost of lower margins. The same orthodoxies suggest to executives appraising a strategic plan that they should funnel cash to almost-leader brands even if it means they must divest small-share brands to acquire the necessalT funds. MBA students are taught that profit follows market share in the general case. But without an understanding of why this relationship might occur, it is difficult to know whether a specific company or product category exemplifies or defies the conventional wisdom. Therefore, it seems useful to ti T to analyze the rationale for expecting long-term payoff for a leading brand versus a small-share brand, and to clarify why the rich brand gets richer. At least three explanations can be offered, one recently published and two presented here to buttress the same viewpoint: large-share brands are the likeliest to gain share. All three arguments for this generalization rest on observations of a snowball effect from category leadership: leading brands have a disproportionate advantage simply because they are leading brands. One explanation for this snowball effect is presented by Ehrenberg et al. (1990), who conclude that small-share brands face double jeopardy. Not only do fewer buyers select them, but they select them less often than the buyers of dominant brands select their favorites. The authors explain this p h e n o m e n o n convincingly using both statistics and psychology. They point out that small-share brands suffer disproportionately from the consumer propensity to buy more than one brand within a product ca tegoiT-to split their patronage among multiple brands. The problem for the small-share brand is that splitters whose first choice is a follower brand are disproportionately likely to divide their purchases between that follower brand and a leader, because they are aware of the leader. By contrast, splitters who favor a leading brand are less aware of small brands. Thus, their splitting represents a smaller proportion of total purchases going to other than their favorite brand. There is a greater likelihood that when they do stray beyond their favorite, they will select another large-share brand, not a smallshare brand. However, it is an oversimplification to assert that rich brands get richer only because of the behavior of splitters. The case becomes even stronger when two additional arguments are taken into account: some leading brands are chosen because of a wish to do what others do, and some are chosen because of a wish to conform to the tastes of as large a group as possible. These two issues will be discussed in turn.