If everything is equally prominent, then nothing is prominent. It is the job of the designer to advise the user and guide them to the most important or most promising choices (while ensuring their freedom to go anywhere they please).

On today's Web, the most common mistake is to make everything too prominent: over-use of colors, animation, blinking, and graphics. Every element of the page screams "look at me" (while all the other design elements scream "no, look at me "). When everything is emphasized, nothing is emphasized.

But it's just as bad to make everything equally bland. Or to alphabetize items instead of sorting by importance.

Here are some ways of using prioritization to guide users:

  • Editorially select the most important stories or items. Give them bigger headlines or more prominent placement. Old principle which newspapers have used for more than a hundred years.
  • Use sales statistics to discover the best-selling products and place them on top of search listings. By definition, most customers will be looking for the best-sellers, so it is user-hostile to bury them in a search listing that is organized by some impenetrable information retrieval algorithm (or worse: sorted by SKU numbers or other internal attributes that don't matter to users). Look at the search results for "Palm" on Buy.com and you will see three best-sellers on top, followed by about 60 other products (other than good prioritization, Buy.com has a miserable search results page: hard to scan; weird abbreviations and symbols).
  • Use server traffic to track areas of the site that are seeing unusually strong activity and place links to these areas on the home page: not only will you save users clicks, but it's also a way of making people aware of the current buzz. The Motley Fool does so to good effect by keeping abreast of the activity of its many discussion boards and placing references to humming ones on a "hot topics" page that is linked from the home page (and summarized right on the home page).
  • Use reputation management to predict who will write the best contributions: if somebody was highly rated in the past, then their new material deserves featured placement. Epinions has reputation data that identifies the most trusted reviewers, and it gives high prominence to these writers' postings even before they have been rated by anybody.
  • Simply highlight the most popular items in a list that is sorted by another criterion.
    • As a more advanced idea, show multiple levels of importance by different levels of highlighting or different icons. This is usually overkill and too much work for users to understand.
  • On slowly changing pages, mark new items with a little "new" glyph. This is not necessary on pages that change all the time (say, newspaper home pages) since the assumption is that most items will be new on such pages.

There are two main types of prioritization:

  1. In lists of items, make sure the ones the user is most likely to want come out on top or are made to stand out.
  2. Content that is deep within the site sometimes needs to be brought out and featured at higher levels to make users understand what's new or hot.

The goal is to give users more of what they need. And easier access to what they need. This is not always the same as giving people what they want: Customization does allow users to set their own priorities; thus it is one way to identify content that should be highlighted or featured. But the user's own choices are insufficient as the only basis for interface prioritization. The other mechanisms I have discussed must be employed as well to guide users to things they didn't know they needed.