My post on Arnold Bread inspired a surprising amount of blogosphere commentary, as well as a long series of comments here (many of them containing a puzzling sense of affront – no, I do not want to regulate bread varieties!) Here’s Tyler Cowen, Jim Manzi, Don Boudreaux, and Russ Roberts.
The wonder of hive mind is that it often provides you with experts who can correct your misconceptions, but the verdict in this particular case was mixed. In the Marginal Revolution thread, people who claimed professional expertise both confirmed and debunked my hypothesis.
One frequent objection was that manufacturers pay retailers for their shelf space. This forced me to abandon my much-cherished lazy hypothesizing blogger posture and actually do some googling. This FTC study (PDF) seems to indicate a complicated picture. Slotting allowances, as this practice is called, are not a universal practice. Roughly half of their surveyed retailers reported receiving them in the bread category. Moreover, typically these fees are paid when a new product is introduced; pay-to-stay fees are rare (although they do exist.) It seems fairly apparent that the strategy of amplifying your shelf space with multiple product lines could easily co-exist with slotting allowances, and might even be enhanced by it.
Also, I think some people seemed to think that my observation must have meant that the retailers would have to be chumps in order for this strategy to work. Emphatically, no. They can be rational players and still give a proliferating product line additional space. Say there is a drugstore with space for six different toothpastes. Big Toothpaste brand is the market leader with 50%, and then there are five other toothpastes with 10% market share. What if Big Toothpaste introduces a product that is under its brand but slightly different? Maybe now each of their splinter brands gets 25%. However, they have driven off a competitor, so the loser’s market share is divided by the remaining players. And now Big Toothpaste has 1/3 of the shelf space, instead of 1/6. Continue this process, and you realize why Colgate has 52 different brands. The retailer is not being irrational; they are serving their customers in the best way they know how. Of course, this strategy has its limitations, and is probably only available to a market leader, but it’s a viable strategy and market tastes have at best a limited effect on why it is a successful strategy. And of course there are multiple drivers. Observing a dynamic does not mean denying that all other dynamics exist.
By the way… I am not the slightest bit outraged by this. Just interested.