Persuasive Advertising - J. Scott Armstrong [27]
Finally, websites can be designed so that customers can specify the features that they are looking for and then receive a list of the products that match their specifications. This makes it easy to narrow a large number of choices to a manageable set.
Evidence on making choices from many multi-dimensional options
Shoppers in an upscale grocery store in California were offered taste tests of jam. In some cases this was done when they passed a table offering 24 flavors, while in other cases there were only six flavors. Of those passing the 24-flavor table, 60 percent stopped for a sample versus only 40 percent at the six-flavor table. This is consistent with previous research that people report a preference for variety. However, at the table with many choices, fewer than 3 percent made a purchase. In contrast, 30 percent of those visiting the six-flavor table made a purchase. Thus, the limited-choice table sold nearly seven times more jam (Iyengar and Lepper 2000). Interestingly, however, an attempt to replicate the “jam study” did not hold up (Scheibehenne, Greifeneder and Todd 2010).
A number of other studies provided findings similar to the jam study. In one, employee participation in pension plans was found to be higher when the number of options was smaller. For example, when two funds were offered, 75 percent participated; however, when 59 funds were offered, only 60 percent participated (Iyengar, Huberman, and Jiang 2004). This study assumes that people are better off by participating and that there is no benefit in deferring a choice.
In the previously-mentioned lab experiment by Gourville and Soman (2005), when subjects were presented with a single oven, 53 percent said they would purchase it. When given a choice of five ovens that varied on different dimensions, such as power, on-line help, and warranty, the intentions to purchase fell to 40 percent.
Evidence on the importance of categories is provided by the following experiments. While the context is retail shopping, we might generalize to organizing choices in print ads, point-of-purchase ads, and Internet ads:
Poorly informed customers were more satisfied when the products were put into categories. Mogilner, Rudnick, and Iyengar (2008) conducted surveys of shoppers in the magazine aisles in ten branches of a supermarket chain store. The stores differed in the number of magazine options and in the categories used to display these options. Irrespective of the number of magazine options, when there were more categories, customers thought there was more variety and they were more satisfied with their choices.
Then, in a lab experiment, participants were asked to select from among an assortment where the number (144) and placement of magazines remained constant. But the number of categories differed:
1) three categories (“Men’s,” “Women’s,” and “General interest”) of 48 magazines each
2) 18 categories (e.g., amongst the men’s magazines, there were categories such as “Auto,” “Hunting and fishing,” and “Sports”). Thus, there were eight items in each category.
The categories had no effect on the satisfaction of informed shoppers, who presumably had no trouble finding what they were looking for. However, the uninformed shoppers thought there was more variety and they were more satisfied with their choices when there were more categories.
The researchers followed this with a field study in which café customers ordered coffee. The customers were randomly assigned—depending on what table they were sitting at—to one of four treatments. Each customer was offered the same 50 coffee flavors, always presented in the same order. The control group received no guidance, just a list of the 50 flavors. Customers in the three experimental groups were presented with menus with the coffee flavors grouped into ten categories of five flavors each. The experimental groups were given one of the following three kinds of categories:
1) descriptive categories (e.g., “dark roast,” “mild,” “sweet”)
2) uninformative