Indexes Impact Sales

Jump to: Authors Know The Value Of Indexes |  How Indexes Impact Sales |  Automated, Computer-Generated Indexes

“If you don’t find it in the Index, look very carefully through the entire catalog.” From Consumer’s Guide. Sears, Roebuck and Co., 1897.

If you decide against having a professional indexer write the index for your document, consider adding a note like the one above to your index.

It’s a shame that some authors pour themselves into painstakingly writing a text, and then they shoot themselves in the foot by excluding an index. Or worse – by including a less-than-useful index.

Poorly written indexes give readers a false sense of security. They turn to something that looks like an index and they assume it covers all issues in the text. But upon using it, they’ll be frustrated by its inconsistencies, awkward phrasing, clumsy main headings and subheadings, unusual or nonexistent conceptual relationships, and lack of cross-references. (Just to name a few of the most common problems in poorly-written indexes. Most people who use indexes have had this experience many times.

Authors Know The Value Of Indexes

David Holzgang, a successful book author, wrote an article for TidBits, in which he explains the importance of a professionally-written index:
The index for a book is one of the most important features that the book has. As a reader, I find a good index makes using a book a pleasure and a poor index makes finding anything a real chore. As a result, I usually ask publishers to let me pick a professional indexer and pay for the index out of my royalty account. This gives me some control over the index quality and ensures that I end up with an index that contributes to the book… Generally, in my experience, the worst indices are those prepared by authors. Indexing is a specialized skill, and deserves respect. (And, if you think authors work under deadline pressures, consider the indexer who generally has no more than a few days to index the book completely.)
Lillian R. Rodberg wrote “The Index as Marketing Tool” for the Textbook Authors Association about how indexes impact sales:
If you expect your publication to be used after the initial reading, you must invest in a well-designed index tailored not only to your publication, but to your users’ needs. Your information does little good if you don’t provide easy access to it. Publishers of computer and technical manuals must remember that because there is usually no cover-to-cover reading of your publications, your indexes will determine how useful readers find your documents.

How Indexes Impact Sales

Good indexes impact sales.

  • knows that potential buyers do look at indexes when deciding which of several similar books to purchase, so they now include indexes in the “Look Inside” feature.
  • Professors and academic teams evaluate indexes when deciding which books will be used as textbooks. Imagine the sales lost when books alienate this particular audience!
  • Librarians making decisions about what books to purchase say that indexes are one of the most important criteria they factor into their decisions. (I’ve talked to many people who check out books from the library before deciding whether to purchase their own copy.)
  • Book reviews often mention indexes. Selected reviews spanning a variety of disciplines are reprinted in The Indexer, the international journal of indexing.

Keep these factors in mind as you work on locating indexers with the expertise you’re after.

Automated, Computer-Generated Indexes

Concordance (a.k.a. index-generation) features of word processing and desktop publishing software does not replace the need for professional indexers any more than spell-check features replace the need for professional editors. Generating a list of words does not create a useful index.

The valuable content of indexes must be written by an intelligent human. Using these software features still means someone will have to do a lot of work building hierarchical structures and access points (like cross-references and double-posts).

One aspect of language that software can’t pick up on is inference. For example, when texts discuss meal plans for dogs, software will generate entries for exact words used, (“meal plans” and “dogs”) but it can’t understand that people looking for information on canine nutrition and diet would find that passage useful. Humans can see this, of course, and skilled indexers would create useful cross-references or double-posts from related or synonymous terms, perhaps from “canine” to “dogs,” and from “nutrition” and “diet” to “meal plans”. The terms they create would depend on the audience and the context – another subtle detail that software can’t replicate.

This detail would becomes obvious when, for example, a human looks at a passage and identifies a 15 page passage about dog nutrition, but a generated concordance would, instead, list individual terms that appear on individual pages in that passage.

Human written index:

canine nutrition, 90-105. See also individual nutrients.

nutritional needs of dogs, 90-105. See also individual nutrients.

Concordance generated by software:

dog, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 102, 103

dogs, 90, 91, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105

diet, 90, 94, 97, 99, 100, 105

diets, 91, 92, 93, 98, 101, 104

meal plans, 95, 96

nutrients, 100, 101, 102

vitamins, 102, 103

Software also can’t create hierarchical structures that human readers expect to find in indexes. For example, a business text might discuss voicemail, landlines, cell phones, and text messaging, but might never use the term “communications”. Again, software can’t connect those topics. Human indexers will consider whether readers would be likely to look under a broader topic like “communications” or “telephones” and if they do, they’ll build in a hierarchical structure using subentries for the more narrow topics of “voicemail”, “cell phones”, “landlines”, etc or by using a general cross-reference telling readers to “see specific types of communication“.

For more information about automated and semi-automated indexing check out the following: