Should publishers invest in software for in-house indexers? A case study

I learned to index on the job—and by reading books like Nancy Mulvany’s Indexing Books—when I worked as an in-house editor. I created several indexes using only Microsoft Word, which is perfectly adequate for projects like cookbooks but can be painful to use for more complex projects that require thoughtful and accurate cross-references between topics and a consistent way to combine and split headings during editing.

The year I started indexing, I spent my professional-development allotment on an indexing course, where the instructor showed us how she worked with her indexing software, and I lobbied my supervisor to get a license for our office. Fortunately, I didn’t have to argue hard—she recognized that the software would pay for itself over a handful of projects. I know of other publishing houses that have chosen to stick with a Word workflow and haven’t bought the software. On one hand, I understand—the price tag of ~US$550 may not seem worth it if they’re only preparing a few indexes in house each year. On the other hand, they’re paying for editing time that wouldn’t otherwise be necessary.

Software won’t help you pick out topics to index—that part still requires a human brain (for now)—but it will reduce the cognitive load of indexing by automating alphabetization, certain aspects of formatting and punctuation, and the order of the locators. Most indexing programs also have time-saving features like autocomplete and error checking for blind cross-references and orphaned subheadings. The final index obviously still needs to be edited, but if it’s prepared using software, the editor can focus on content and organization rather than on nitpicky (but essential) details like alphabetization.

Recently I had to edit an index that a publisher created in house—without indexing software. I thought I’d use it as a case study to quantify how much time using software would save. I won’t comment on other issues of quality like term selection or accuracy and comprehensiveness of the locators but will focus on problems that software would have obviated.

The index was just under 5,000 words and was for a 300-page historical atlas.

I spent 6 hours and 57 minutes editing and proofreading. This was probably a little longer than I would devote to most projects, but this book had a peculiar design workflow.

Of that time, I spent 50 minutes checking alphabetization and found several inconsistencies in how characters like ampersands were treated. I mention these inconsistencies not as a criticism of the indexer but as a justification for why this check was necessary.

The subheadings of a particular heading were not properly alphabetized at all, and when I looked into it, I discovered that the line breaks between subheadings were manual ones, so Microsoft Word’s sort feature didn’t consider them separate paragraphs. This problem wouldn’t arise with indexing software.

I devoted 26 minutes to checking the locator order. In general, this aspect of the index was well done: I found only one error. But again, I wouldn’t have had to do as close a read for an index compiled with software.

I spent 10 minutes checking formatting of cross-references and confirming that the pointers matched the targets (and I found a couple of errors there). I also noticed that the commas in the document weren’t consistently formatted after italicized or bolded text, another problem that wouldn’t usually arise with an index creating using software.

I spent 30 minutes double-checking alphabetization and locator order during the proofreading stage and found a few changes I’d missed making.

So, 117 of 417 minutes (a conservative estimate—because the workflow was unusual, I haven’t included the time it took me to implement the changes in the files) were spent on checking issues or fixing problems that software would have taken care of. If my editing fee had been hourly, the publisher would essentially be paying a 28% premium for my work. At that rate, the software would pay for itself in 6–8 indexes. I haven’t even considered the time that indexing software would have saved the indexer—at least as much as it would have saved me—in which case the software would have been paid off after 3 or 4 indexes. (And I’m still using the same version of the software I bought 6 years ago.)

This is just one data point, but I hope it shows the value of indexing software, even for small presses, if they do any indexing in house. In the indexing course I teach, students have a week to explore demo versions of three industry-standard programs and use them to build a simple index, so the learning curve is not that steep. In addition to saving editing time and cost, it also eliminates the frustration while editing of knowing that the process could have been a lot simpler.

Leave a Reply

Your email address will not be published. Required fields are marked *