The Missing Element—Information Quality

Home/Publications/CIDM eNews/Information Management News 02.08/The Missing Element—Information Quality

Kent Taylor, acrolinx North America

When we consider the most important elements of a successful single-source strategy, we often miss one critical element. This is the same element that practically everyone else misses in their rush to structured writing, content management, single-sourcing, and ‘doing more with less.’ It is the same one that I also missed for most of my many years in Tech Pubs–because there was no effective technology to deal with it.

That missing element is the quality of the information itself – the pesky stuff between the tags, the only thing the readers actually see. It’s difficult to quantify, and I don’t think it can be defined as a stand-alone step; it’s something that spans a single-source and minimalist strategy, that JoAnn Hackos present in her Director’s Column in the February 2008 edition of Best Practices. So, I’ll try to weave it into JoAnn’s framework.

  • First—All content must be planned as a whole. The plan should include standards for style, terminology, readability, and translatability, with agreed-upon threshold levels for conformance to these standards.
  • Second—all content must directly serve the needs of users for information and action. An often overlooked need of users is their need to be able to easily find, read, comprehend, and digest the information that is available to them. Today, many companies’ product information goes to an increasingly large population of non-native speaking readers, whose vocabularies and knowledge of English sentence structure is nowhere near that of the average technical writer. And, sadly, only 20% of Americans read at a 10th grade level or higher. Tech Writers, on the other hand, read at college or postgraduate levels, and have large vocabularies. (Draw your own conclusion …)

A large and growing user need is to provide the information in their native language. Globalization is here to stay, and translation needs will continue to grow exponentially for the foreseeable future. Translators, both human and machine, have needs that generally go unaddressed by the authoring community, resulting in translation times and costs that may be 15% to 50% higher than they need to be.

  • Third—all content must be consistently structured. All content should also follow consistent style, and grammar standards, use consistent terminology, and consistently avoid known translation problems. The use of XML authoring tools helps to enforce consistent structure; commercially available information quality tools can help to enforce conformance to style, grammar, terminology, and translatability standards.
  • Fourth—all content must follow the rules of an XML-based content structure. All content should also follow the rules contained in corporate style, grammar, terminology, and translatability standards. And just like XML-based content structure can be checked by software before going to production, the properly structured content should be checked against corporate quality standards by software before going to production.

‘Everyone’ knows about structured writing, XML authoring, content management, and single-sourcing tools, but few are aware of the AI- and NLP-based Information Quality Management tools that are available today. These tools can analyze and evaluate text quality more effectively and orders of magnitude faster than the best human copy editors. And a growing number of forward-looking companies are building these tools into their Information supply chain processes to realize substantial improvements in information quality and readability, with the added bonus of 10% to 25% reductions in time-to-global-market and information supply chain costs.

You know that I agree with your basic premise. I believe technical communication processes are in transition from an individual artisan model, to a team-based mass production or manufacturing model. But I also believe this model must ultimately include Quality Assurance and Quality Control tools and processes, to ensure that the interchangeable components actually fit and work together.

Bottom line: This is not your father’s tech pubs process, and it won’t be welcomed by everyone. But it will eventually impact everyone. The die is cast, and the movement is, in my opinion, unstoppable.

I think it’s about time for the tech pubs community to be aware of and start a dialog about these information quality issues and how to deal with them.

To read JoAnn Hackos’s original article in Best Practices, consider subscribing today if you are not already a CIDM member.