A-Z Indexes to Enhance Site Searching
Published on January 5, 2005
Facilitate searching, not just navigation
An important part of an information architect’s job is to make it easier for users of a Web site or intranet to find the information they want. Usually the focus is on site navigation—the site’s structural design, hierarchy, page titles and labels, menu design, site map, and so on.
Another way to address making information on a Web site easy to find is through search functionality. What’s the difference? Navigation means finding one’s way around and learning the layout of the site. Searching means finding a desired bit of information as efficiently as possible. A good site should support the search needs of users, not just the navigation needs.
When we think of searching in the context of the Web, the idea of search engines immediately comes to mind. Search engines, a practical way to find information on the entire World Wide Web, are increasingly being added to individual Web sites to allow users to search a site. However, results tend to be less than satisfactory.
Drawbacks of search engines
Search engines only pick up exact words or phrases. If a user enters a synonym, singular instead of plural, a spelled-out form instead of an acronym, a misspelled word, or merely a concept with words that never appear in the text, appropriate pages may be missed. Searching the entire Web, missed pages are usually not a problem since so many results are retrieved. But on an individual Web site, it is essential that all relevant pages be returned.
Search engines pick up pages that contain a specific search phrase, even if just in passing or out of context. The page could be about an entirely different subject. This isn’t a huge problem when searching the entire Internet because major commercial search engines have developed complicated ranking systems based on meta tags, keyword frequency, links, etc.
Of course, a site search engine can be customized to search only keyword meta tags as long as keywords are carefully created for each page. If you are going to go to the trouble of creating keywords for each page, you may was well create a manual index for the Web site. This option has several distinct advantages.
What is an A-Z index?
As an “index” can have different meanings, as can “site index.” According to the National Information Standards Organization TR-02-1997 standard, an index is “A systematic guide designed to indicate topics or features of documents in order to facilitate retrieval of documents or parts of documents.” NISO classifies indexes as displayed and non-displayed, and further explains that a displayed index has syntax for combining terms in headings and a systematic ordering of headings. The most common systematic ordering is alphabetical, and being displayed means that it can be browsed. For Web sites or intranets, this type of index, to distinguish it from other indexes, is often called an A-Z index.
On a Web site or intranet each of the alphabetically arranged entries or subentries is hyperlinked to the page or to an anchor within a page to where the topic is discussed. Since an alphabetical index can be quite long, it is often divided into pages for each letter of the alphabet. Typically, each letter is linked at the top of the page allow a jump to the start of that letter’s section of the index.
Here is an example of a Web A-Z index:
Advantages of Web A-Z indexes
Unlike typical search engines, A-Z indexes created by expert indexers point to all substantial information about a topic. Nothing is missed, and extraneous pages are not retrieved. Additional advantages include the following:
A-Z indexes are the user-friendliest way to search. Users are already familiar with how to use browsable alphabetical indexes from books. The user only has to pick a term off a list and not have to think up something to type in a search box. A user can browse a list of alphabetical topics with a sense of security that each topic will yield a result and the resulting page will have more than just a cursory mention of the topic.
The browsable nature of the index can reveal other topics of interest to the user. The user might find additional information beyond the original search objective. This may increase user satisfaction with the site and its content and the user may stay longer or visit the site more often.
Index entries can link to precise points within a Web page through the use of named anchor links. The user does not have to scroll through the retrieved page to find the desired information.
An A-Z index can enhance the search engine optimization ranking of the site. A large number of new internal links created and the words within the linked text are high-quality keywords.
What Web sites are best suited for A-Z indexes?
The first question to ask when deciding whether to create an A-Z index is whether the site is one that users want to search—not merely to explore but to quickly find a specific piece of information. Intranets instantly come to mind, and are the most appropriate type of site for an A-Z index.
Other kinds of sites for which users especially appreciate indexes include:
- Colleges and schools
- Research institutions
- State and federal government agencies
- National associations
- Other sites acting as a comprehensive reference source for a particular subject
Media sites or e-zines, which contain archives of articles, images, or program synopses, can certainly benefit from A-Z indexes. However, since new pages are frequently added, an index as an interface to a dynamic database for a set of pages is the most practical solution.
An example is the index of the Montague Institute Review.
Corporate external Web sites that can benefit from A-Z indexes are those that draw repeat visitors, such as banks or insurance companies visited by their clients. Corporate external Web sites that aim merely to give a Web marketing presence to a company, and not to give detailed information about numerous products, generally do not need indexes. Such sites are designed for one-time visitors to navigate and explore, rather than to search.
A-Z indexes work best with medium-sized Web sites or intranets of between 30 to 300 pages. If a site is extremely large, it becomes impractical to create and maintain an A-Z index for it. Although sites of several hundreds of pages can be manually indexed, just as books of several hundreds of pages are manually indexed, the dynamic nature of Web pages makes such large site indexes more challenging to maintain. For large sites, it may be more practical to write an A-Z index for only the most static part of the site or to incorporate the A-Z index into a dynamic database from which documents can be retrieved.
Even sites as small as 20 pages can be served well by an A-Z index. Although less necessary on a small site, the addition of an A-Z index can enhance the site’s professional appearance. An index should be avoided if the site is so small and well organized that most pages are accessible from the main menu.
A-Z indexes are more practical on sites that are not constantly changing. Specific content within pages can be updated without affecting the index, as long as the topic of the page or page section remains the same. If pages are added and deleted in an unsystematic manner, then more work is required to keep the index maintained. If pages that get added tend to be of a consistent type, then the original indexer can write indexing policy guidelines that the Webmaster can follow. So, who creates the index?
Skills needed to write an index
Writing an A-Z index, with its two levels of entries and subentries, inverted terms, cross-references, etc., is a professional skill that requires training. It is normal for no one on a Web team to be skilled in indexing. Information architects with training in library and information sciences, however, may have taken a course in indexing, and for them a review of the course materials should suffice.
Information professionals who have worked only on categories and taxonomies, however, should realize that A-Z indexes are not exactly the same. For those with at least some background in indexing or taxonomies, an additional course is probably not needed but a review of any of these books is recommended:
- Indexing Books, by Nancy C. Mulvany
- The Art of Indexing, by Larry S. Bonura
- Indexing Concepts and Methods, by Harold Borko and Charles L. Bernier
- Facing the Text: Content and Structure in Book Indexing, by Do Mi Stauber
- Handbook of Indexing Techniques: A Guide for Beginning Indexers, by Linda K. Fetters
- Indexing: A Nuts-and-Bolts Guide for Technical Writers, by Kurt Ament
Indexing courses and books tend to emphasize back-of-the-book indexing. For Web site indexing, the basic skills are the same. It is important to realize that in Web site indexes, each entry or subentry can be linked to only one page, unlike a book index where the entry or subentry may have multiple page numbers listed afterwards. Therefore, creative solutions may be required. It is highly recommended to review existing Web A-Z indexes to get an idea how they are done.
Unless someone on the Web team already has some training in indexing which further reading can enhance, it is preferable to contract out the A-Z index to a professional. Although writing an index requires a trained skill, indexers do not expect to be compensated as highly as information architects. Professional indexers tend to be independent contractors, and most belong to their respective national professional associations, which maintain directories of registered members. The following indexer directory databases can be searched by index type, such as Web/HTML.
- American Society of Indexers
- Indexing and Abstracting Society of Canada
- Society of Indexers (UK)
- Australian Society of Indexers
Some local chapters of the American Society of Indexers also maintain indexer directories, listing indexers who may not be registered in the national indexer locator directory.
Web indexing tools
Automatic generation utilities
You may come across inexpensive or free software tools or extensions that automatically generate Web site indexes or site maps. Generally, these merely create an alphabetical list of titles of the site’s pages. That would be like alphabetically arranging the entries of a book’s table of contents and then putting it in the back of the book and calling it an index, which is unheard of. Such low standards for Web site A-Z indexes threaten to lower the expectations and opinions of A-Z indexes in general.
Index editing software plus HTML-conversion tool
Most professional freelance indexers use one of three commercial indexing software packages: Cindex, Macrex, or SKY Index. These programs aid in the alphabetizing of entries and subentries, creation of cross-references, indenting of subentries, formatting, and other index editing tasks. To convert an index created in one of these programs to an HTML document, use a utility called HTML/Prep. An index must first be created, and URLs for links instead of page numbers need to be manually pasted. HTML/Prep then converts the compiled index file, which has been saved as text, into an HTML file, preserving the index format style and adding hyperlinked letters of the alphabet along the top of the page.
HTML index editing software
If you feel comfortable with the techniques of indexing but do not own a dedicated indexing program such as Cindex, Macrex, or SKY Index, the most comprehensive tool for back-of-the-book style indexes is HTML Indexer. With the site’s files on the local computer, HTML Indexer automatically creates an editable entry for each page or page plus named anchor, with the URL already retained. Indented subentries and cross-references are supported, but it lacks more sophisticated editing features.
Database management software
For database indexing, any database management software can be used, but you need to invest time and effort to design the database for the purposes of the index. Unfortunately, there are few database packages specialized for indexing. Authex is designed for periodical indexing, but its developer is no longer supporting this DOS program. A better solution might be ProCite. Although it is designed for maintaining bibliographies, professional have used ProCite for periodical indexing.
A-Z indexes are most suitable for intranets or sites with repeat visitors, and for sites of a medium size that are not changing too rapidly.
Cheap or free site search engines and “site index” generation tools can quickly set up some form of searching for many sites but for a highly effective search and increased user satisfaction, a professionally written A-Z index is still the best option. An A-Z index as an interface for a database is a possible solution for sites with pages added or removed frequently.
If human-crafted indexes have been effective in helping readers find information in millions of non-fiction books and manuals, then they ought to be useful for many content-rich Web sites.
Heather Hedden is an information taxonomist with Viziant Corporation in Boston, MA. She also does consulting through Hedden Information Management and teaches continuing education workshops through Simmons College Graduate School of Library and Information Science.