To explore the indexing profession, interested high school students should visit libraries to read and evaluate indexes of all kinds and develop a solid general knowledge of many topics. Keep up with the rules of grammar and spelling because the career depends on accuracy. Students also should read some basic books on the practice and theory of indexing, such as Indexing Books (Chicago: University of Chicago Press, 2009) or Handbook of Indexing Techniques: A Guide for Beginning Indexers (Medford, N.J.: Information Today, Inc., 2013). The American Society for Indexing offers a variety of information on indexing and getting started in the field at https://www.asindexing.org/about-indexing. Finally, talk to an indexer about his or her career. 

Indexers are professionals who truly enjoy words and text. There are several types of common indexes. The most familiar is the back-of-book index. Back-of-book indexes contain references to information in only one volume. Most nonfiction, single-volume texts include this sort of index. Multivolume indexes contain references to information in more than one volume. The page references in a multivolume index must indicate clearly both the volume number and the page number of the cited information. Most encyclopedias include multivolume indexes. Magazines and newspapers also have indexes. These periodical indexes are published separately, at regular intervals throughout the year, and are extremely helpful to researchers.

Indexers may also develop a topic specialty, such as law, medicine, finance, or science and technology.

A more recent development in indexing is the online index. Online indexes help users locate specific information from within a large database. Online indexes differ from a simple search function in that an indexer has created a translational thesaurus. When a user inputs a term that actually does not exist in the database, the online index will translate the term to a synonym that does exist so that the user may access the needed information.

Though their scope and purposes vary widely, all indexes have certain features in common. Every index must be organized according to a useful system. Most indexes are alphabetical, though in some specialized cases they may be chronological or numerical. The index to a history text, for instance, might be in chronological order. The two most commonly used alphabetical filing systems are the word-by-word arrangement, under which New York would precede Newark, and the letter-by-letter arrangement, under which New York would follow Newark.

All indexes must contain index terms, called headings, and page numbers or other locators. Most indexes also contain subheadings that help users narrow their search for information. Under the main heading "George Washington," for example, an indexer might use subheadings to separate references to the Revolutionary War from those to Washington's presidency. An index also may include cross-references to other pertinent headings or indicate the presence of illustrations, charts, and bibliographies.

Whether one creates an index on three-by-five index cards or with the help of a software program, the mental process is the same. The indexer first must read and understand the primary information in the text. Only then can the indexer begin to identify key terms and concepts. The second phase in compiling an index is called tracing—marking terms or concepts. Choosing appropriate headings is often the most challenging aspect of an indexer's job. Subjects must be indexed not only under the terms used in the text, but also under the terms that may occur to the reader. Since the indexer's first obligation is to help the reader find information, the best indexers ask themselves, "Where would the reader look?"

After tracing, the indexer begins to compile the headings and page references. Entries with many page citations must be divided further by subheadings. The final step in creating an index is editing. The indexer must view the index as a whole in order to polish the organization, delete trivial references, and add appropriate subheadings.

While indexers may organize information by keywords or concepts, the most useful indexes usually combine both systems. Keyword compilation is indiscriminate and is of limited usefulness to the reader. Keyword lists include every instance of a term and usually fail to make connections between synonymous or related terms. Computer programs that promise automated indexing are actually capable only of compiling such keyword concordances. In conceptual indexing, on the other hand, the indexer is not bound to standardized terminology, but recognizes synonymous or related information and disregards trivial references. Even the most sophisticated computer program is incapable of creating an adequate conceptual index.