Why using special characters as bullets is a bad idea

  1. It dilutes the true meaning of the characters when you use them out of context.
  2. Not all fonts contain these characters so some users may see boxes in place of them.
  3. Screen readers for the blind and translation software may have trouble with them.
  4. The proper way to embed them into HTML is with an escape sequence such as » for guillemets. Sooner or later you're going to let somebody else work on your website who doesn't understand this and they're going to copy and paste text containing them into a new page. When they paste the character, it's likely to write it into the document as raw data instead of the escape sequence. This may cause the character to display incorrectly for some people.
  5. In many cases, these characters are used to replace bullets in a list. Usually this means the author is no longer using <ul> and <li> tags. These tags actually have semantic meaning. They group the elements together into a list. A bunch of text prefaced by &raquo;'s and followed by <br> tags does not have the same meaning though it may look similar on the screen.
  6. In the case of a website such as a CafePress Premium shop, the characters can't be easily removed or styled by the shopkeeper. Unordered lists are highly "stylable". Suckerfish dropdowns from A List Apart are one of the more dramatic examples but unordered lists can be styled into all sorts of different navigational elements. For example, the collapsible tree and tabs on Spinlife.com's review page are both unordered lists (<ul> and <li>). The top and side navagation on Wheelchair.net are also unordered lists.

If a custom bullet must be used that isn't supplied by the browser, it should be done as an image, not a special character. In addition, if at all possible, it should be added via CSS and not in the HTML source.