Traditional Use of Indexes
Have you ever rented a car, and then found yourself searching the car manual for instructions on, say, how to release the hood? You probably went straight to the car manual’s index, rather than reading through the table of contents.
Have you ever searched a cookbook for a carrot soup recipe? If so, did you look through soup section of the table of contents, or did you jump to the index in the back? If you turned to the entry for “carrot”, the index might have directed you to the entry for “soup.”
Why did you use the index rather than the table of contents? Because the table of contents is there to give a sequential overview of the topics in the book, whereas the index directs you to topics related to a specific subject you need. The index is the quickest and easiest way to find specific information. So everyone agrees that printed material can’t do without indexes.
However, many documentation teams ask whether indexes in non-printed materials still have value. Keep reading to get each side of the debate.
Why Indexes are Still Useful in WebHelp
Reasons for keeping indexes include:
Indexes allow for more fine-tuned retrieval, with possible sub-index entries, and “see also” entries. Sometimes a user might even browse the index and then happen upon the appropriate term.
2. Author Knows Best
A human can more accurately determine which terms are related to the topic than the best search algorithm would be able to. Although search algorithms are improving every day, and now include fuzzy search, synonym search, and even spelling-mistake search (gotta love that one!), they still cannot compete with the author who knows exactly which terms are relevant to the topic at hand.
3. Index Terms Are Already There
Assuming that the content is single-sourced to both printed and virtual documentation formats, someone has already gone to the trouble to index each topic. Why shouldn’t that effort be leveraged elsewhere? Adding the index seems to be a “no-brainer”.
4. WebHelp Convention
Users have come to expect the index, along with TOC and Search. Removing the index would be doing away with an anticipated point of navigation.
Why Indexes in WebHelp are Nearly Obsolete
Reasons for deprecating indexes include:
1. Search Does It All
Search and context sensitivity allow you to find any topic by keyword, so that the advantage of looking through the index is marginal.
2. Nobody Uses the Index
It is widely known that most users go straight to search without ever glancing at the index. In any case, index terms that were added to the topics can be used by search engines to give weight to the search results, removing the need for a separate index.
3. Real Estate
The index adds clutter to an already full screen, which is especially problematic as screens keep getting smaller and users have less and less patience for distractions on the screen.
4. CJK Sorting
Sorting in Chinese, Japanese, and Korean is quite complex for reasons which are beyond the scope of this article. Antenna House has an index sorting module available for these languages published to PDF, but no such code exists at this time for WebHelp. The only solution at this time is to have the index entries manually sorted by the translation vendor. Is it worth the high cost of sorting these terms if no one uses the index anyway? Or should the localized index terms be haphazardly incorporated to populate the index?
The Bottom-Line Question
Is the fact that users are accustomed to the presence of an index enough of a reason to keep it? And in the final analysis, it is the user who determines how useful or necessary an index is.
What are your thoughts on the value of indexes in WebHelp? Has your team debated this issue? If so, what were your conclusions? Let us know in comments section below.