If one reads the research behind these two claims carefully and interprets the results across the studies, one can come up with what seems to be a reasonable, additional view of actual consumer behavior. Both sets of statistics could be about right, and perhaps the studies should be read as complimentary.
Most consumers are not entirely loyal to a specific brand in a category. Most have a set of acceptable substitute solutions that they chose from based on:
- availability
- price
- other promotions or incentives
- other factors likely to impact the current purchase (“Fred’s mom does not like my brand”)
Traditional grocery lists (perhaps all shopping lists) serve primarily as “reminders” to the consumer triggering an in-store scan and grab for the product of primary choice or select from a largely pre-determined set of substitutes. So, the list while not explicitly calling out products, does evoke a specific customer centric set of product choices for each list entry. The list determines a small set of products, the final choice is made at the shelf based on the conditions or circumstances laid out above.
If I am correct about this, an I see no reason to doubt myself 
, this should critically influence how application and ecommerce platforms and loyalty service providers think about building list capabilities.
Back when I designed the MyWebGrocer online grocery shopping product selection display, I chose alpha by brand/size as the hierarchy to be used. After we had launced and run for a bit I looked at the impact of this display hierarchy on consumer product selection behavior.
I worked with Lisa Selip at Lowes Foods and looked at market share comparisons between online customers and in-store customers on toilet tissue. I picked that category because it tends to be purchased frequently, universally and there is not a huge switching penalty between brands. The result was surprising and probably a bit dismaying for brand managers. The market share for the first products displayed in the category was almost double the share generated on those same items in-store.
My conclusion about these seemingly disconnected studies, is that while a product might be on a consumer’s mental or actual list , loyalty to that selection is generally quite tenuous. Meanwhile influence is and can be exerted at the shelf to move a consumer from a primary selection to a pre-selected substitute. It is much harder to move them to an item not on their evoked list at all. The implications on the designers, developers and advertisers who wish to gain access to that consumer list are far ranging and complex. The rewards for figuring it out are monumental.
