The Costco Effect: Science Says We Choose Less Variety When Buying In Bulk

When you go to a convenience store to grab a few cold drinks and some snacks, you’re probably going to make different shopping choices than you would at the supermarket or warehouse store. And a new study claims that we tend to go for more variety when we’re not buying in bulk — even if the bulk packages offer variety.

A new study [PDF] from an international group of researchers published in the Journal of Consumer Research looks at consumers’ desire to opt for variety when purchasing single products versus purchasing them in bundles.

The first experiment divided test subjects into two groups. The first group was twice asked to choose between Coke or Sprite. Thus, the subject could choose a Coke followed by a Sprite; or a Sprite then a Coke; or make two consecutive choices of the same beverage.

The second group was given one decision to make: Choose between bundles of two sodas — either two Cokes, two Sprites, or a variety pack of one of each.

According to the study, 62.1% of subjects asked to make two separate choices opted to go with one of each drink. Meanwhile, only 34.2% of those presented with the bundled options chose a Sprite/Coke combination.

The researchers also asked subjects if they had a preference for either of these brands and rate their preference on a scale of 0 (for no preference) to 6 (for the highest level of brand preference).

As you’d probably expect, subjects in the bundled-pair group who expressed a preference of 3 or higher always chose their favorite soda. But that brand loyalty didn’t hold true for the other group, where around 25% of people with a strong preference for one soda still opted for variety in their selections.

“[T]he impact of performing two choice acts on variety seeking was such that many participants in the single offering condition chose two different items despite acknowledging on the liking rating scale a much stronger preference for one soft drink over the other,” reads the study. “Therefore, it seems that it is the single offering frame that pushes choosers to seek variety.”

This hypothesis was backed up by a second experiment involving the same setup, but replacing Coke/Sprite with Snickers and Twix candy bars. In this case, 73.6% of participants who made two individual choices selected both snacks, while 50% of those forced to choose a bundle opted for variety.

The candy bar experiment added another consideration. This time, the candy-selecting subjects gave feedback on the thoughts they made when making their choice. Then third-party participants were asked to review those answers and guess whether the participant had chosen two different items or chosen two of the same item.

Regardless of whether the subject made two individual choices or one bundled choice, it was quite easy to correctly determine that they had selected two of the same items. The guessing participants correctly identified 79.4% of cases where subjects doubled up on their candy bar choice. This is likely because subjects gave feedback expressing a particular affection for one brand over the other, or because of a dislike of ingredients in one snack.

The guessers also correctly identified 72.4% of those subjects who chose mixed bundles. Researchers says it’s because they used statements like “I really like both of them so I want one of each” and “I don’t want two of the same.”

But those subjects who opted for variety when faced with two separate choices were significantly more difficult to suss out. The guessers only correctly identified 17.9% of these people.

The confusion lies in statements like “I don’t care for Snickers,” or “I like Twix better than I like Snickers,” which seem to indicate a strong preference, but were actually given by people who chose both candy bars.

“In short, many single offering choosers may have been pushed into seeking variety, and as a result they found it very difficult to explain their choices in hindsight,” reads the study. “They instead just reported their preference for one candy bar over the other. Still another pattern of responses in this condition showed that some participants just focused on the product attributes that crossed their minds, which naturally would leave an outsider puzzled about what they had chosen.”

A third experiment attempted to take away any brand favoritism or issues of hunger/satiation by looking at what subjects would choose to do when presented with varieties in floral bouquets.

In this test, one group was presented with an array of different colored roses and told to put together a bouquet. There were enough of any one kind that the subjects could make monochrome bouquets or choose variety.

Another group was asked to choose between a selection of pre-made bouquets, some monochrome and some with a variety of roses.

And as in the two previous experiments, the majority of people (69.6%) allowed to build their bundle chose variety while a minority (48.3%) of those choosing from pre-made bundles opted for variety.

“The question that emerges then is this: Will consumers still be incorrigible variety seekers if they are asked instead to choose multiple items from bundled offerings?” asks the study. “[W]hen consumers seek variety, perhaps they are driven by the choice process (i.e., the mechanics of choosing) as much as they are by the choice outcome (i.e., the items they end up with).”

The researchers believe that their findings have implications for how retailers manage their businesses.

For example, while many bundles are discounted in order to entice customers to clear out inventory by paying a lower unit price, the researchers believe that bundles could also be used to simply entice people to buy more of the same product even if there is no price incentive.

The results also imply that consumers who shop at smaller markets, where things are less likely to be bundled together, are more likely to choose variety than those who patronize warehouse stores and other retailers where buying in bulk is the norm.

“If this conjecture is correct, then it should be easier, other things being equal, for new competitors to gain a foothold in small rather than in large retail outlets,” conclude the researchers.