This blog looks at how to measure your content and your audience feedback as just another tool. As writers, we are presented with a ton of data about our content. I have found it a challenge to find the measures to focus on when creating and maintaining content.
I also recently read the book, Nabakov’s Favorite Word Is Mauve by Ben Blatt. Blatt uses data to analyze famous writers and famous writing prizes. He discovers, for instance, the prohibition against adverbs does in fact correlates with big lit prizes. Avoid adverbs: win the Nobel.
Blatt’s book lead me to look for other approaches to using data mining and data analysis approaches to understanding my own work and literature in general. I found Jodie Acher and Matthew Jocker’s book The Bestseller Code. This book identified a few surprising insights — emotional language tends to be found in bestsellers and again, adverbs, not so much. Jocker is an academic who works in a new discipline called Digital Humanities, and in a way, I found this was pay dirt for using counting and text. Digital Humanities can best be summed by Franco Moretti’s approach in Distant Reading (in contrast to close reading). Moretti talks about being able to gain insight into hundreds or thousands of books at a time rather than focusing on a singular text. His point is that the 19th century novel tends to be represented by a canon of 200 books, and yet there were more than 20,000 novels published in that century.
I want to look at the issues around and the methods for measuring the words used in texts, the upside and downside to common measures, and will provide a catalog of the sources of raw data about your writing with commonly used tools such as Microsoft Word, Pages, markdown editors such as Ulysses and Visual Studio Code, as well as and commonly used audience measurement tools such as Google AdWords and Amazon’s collection of audience measurement tools.
If I had to state the intent of this effort:
Intent: You can gain valuable insight into your content by measuring your content. The measurement of your writing can be part of your content practice, and sets the foundation for creating tests when validating hypothesis about your writing.
How: Look at things in your writing that can be converted into numbers. For instance, you can look at word count and term frequency.
- Look at things in relation to your writing that can be converted into numbers. You can look at length of time readers spend on your article.
- Create a hypothesis: A long article takes a long time read.
- Validate the hypothesis: Long articles have long time on page.
- Collect data on long articles and short articles.
- Assess the outcome.
I have several underlying agendas with this bog.
- Promote the idea of metrics as merely a tool like a dictionary or pencil.
- All metrics provide insight and are also flawed. They are models and not reality. They conform to the idea of good enough rather than perfection.
- Separate metrics of your writing from other things about your writing, such as how much you invoice, your performance as a writer, or other external business type things.
- You can use measurements of your writing as a robot reader. You can then inflict your drafts on the robot before you bother sending them to an unsuspecting audience.
My name is Matt Briggs. I grew up in the Snoqualmie Valley in the 1970s which is now a bedroom community of Seattle’s tech industry, but when I lived there it was the home of hippies, lumberjacks, bikers, hobos, and Sasquatch sightings. I wrote several books set in rural Washington State published by small presses including The Remains of River Names and Shoot the Buffalo.
I now work for Microsoft as a content developer. I’ve worked as a technical writer for eighteen years. You can find my code on GitHub and various tech articles in the Microsoft Azure documentation at docs.microsoft.com