Alphabets

There’s an old joke: Where does “finish” come before “start”? In the dictionary.

Alphabetizing a list jumbles items. So while alphabetical lists look simple, they’re often hard to use. If users don’t know the correct word for what they’re trying to find, they’re lost. Are you looking for a jacket or a sport coat? Do you want to speak to someone in Marketing or Sales and Marketing? Alphabetical lists work well for indexes of proper nouns—where there’s a “correct” word to describe something—like surnames or countries.

Arranging items by popularity is also problematic since we’re unlikely to know what other people like. If I’m looking for pasta in an online supermarket, then I expect to find capellini, linguini, and spaghetti near each other (because they’re all long and thin). And I’d expect to find an 18oz packet of spaghetti next to a 9oz packet—even if the 18oz packet is twice as popular. A list of popular pasta types would feel as random as an alphabetical list. Look at the shelves in a supermarket, and you won’t see a label saying “long, thin pasta,” but that’s how it’s organized. Uncovering those “hidden” categories of metadata is important in creating simple experiences.

Arranging content by format (words, pictures, videos) is another way of categorizing that looks simple but turns out to be unhelpful in the real world. If you’re reading about Hawaii, you want to see photos and videos then and there. Going back to the start to find videos is just too much work.

There are a few situations I’ve come across where organizing by format makes sense. Examples are conference programs in which some formats, like tutorials, require a different registration process. In other words, some formats were used differently by the participants. But these are exceptions—it’s usually simplest to organize conference information by time.

What matters is arranging items according to relevant indexes. Alphabets make sense for staff names, popularity makes sense for top-100 movies, format makes sense for conferences. Choose indexes relevant to the task.