On his blog, Ajit Jaoker argues that mobile widgets support the long tail (LT). I don’t agree, although Ajit doesn’t elaborate. The LT says that it’s equally easy to access any part of the distribution curve for any ranges of products. The classic is Amazon. They can sell books that are 1,299,999 on the list just as “easily” (not necessarily as economically) as number 1 on the list because they have unlimited “storefront space”, which is all virtual. Borders can’t do this with limited shelf space.

The LT is fine, although when I visit Amazon, I invariably get shown products high in the selling rank, not 1,299,999 on the list. Of course, the teaser products can be tuned using personalization.

Unlimited storefront space works on the Internet only because of search. However, there isn’t an equal cost to access information. Some stuff takes a lot longer to find, unless I know exactly what I want and can be assured of exactness in the search database, such as offered by book titles perhaps (if we know them and can type them properly, which increasingly we don’t because we rely on search to help us avoid having to remember or even know such stuff).

Mobile searching is still very costly for all the obvious usability reasons. Therefore, the “best bet” in terms of presenting products to users is to focus on the head, not the tail of the curve. Personalization is important, perhaps more so for mobile. Here I see there is a lot more innovation possible and required, especially utilising personalization information shared across a number of services and channels, from IPTV, Web and Mobile. It is in these areas of information management that mobile innovators need to focus. The current UI fetish won’t solve many of the problems unique to mobile devices and the mobile context generally.