How to decide what to write about?

I write several articles every year, some here on Medium, some internal to IBM. People whom I mentor often ask me “do you find writing is important for your work and your career?” to which the answer is “absolutely, yes.” Then I share what’s become a standard explanation about how I decide what to write about.

I’d like to share that standard explanation here, for reasons that will make sense by the end of the article.

Noise and signal

Before explaining how I choose what to write about, I think it’s worth facing the question “why would I write about anything?” There’s a general, massive problem of too much mediocre, undifferentiated content in the world, so the starting point is to say “I must not do that. Anything I write must be high-quality, and differentiated.” Quality is a matter of practice and care, and a subject for another article, so I’ll focus on differentiation.

I believe that being a good writer starts with being a good reader, so the first part of deciding what to write about is to simply read a lot. I choose what to read almost entirely by following a set of diverse experts on topics I care about on Twitter as well as reading recommendations from trusted friends and colleagues in the course of working together.

So the first principle in deciding what to write about is “don’t write about something if there’s already an existing resource that covers it well enough.” Over the past twenty years, I’ve collected thousands of links and tagged them to aid retrieval, such that if I’m talking to someone and want to share knowledge, I’ll often navigate to my Pinboard account, and in seconds be able to share something great that someone else wrote.

This is how I end up sharing knowledge probably > 95% of the time and answers the question “how to decide what NOT to write about?” With that out of the way, I’ll now address the unintuitive process I’ve developed for deciding what to write about.

Discovering what you know, that others don’t

Expertise is a funny topic because on one hand, everyone has some areas of expertise but on the other hand, most people don’t understand how expertise works [1]. Because of this dynamic, and because expertise grows slowly over time, most people lack self-awareness about areas of deep expertise.

But simply having expertise in a topic doesn’t immediately illuminate what is worth writing about. For instance, I know a fair bit about “DevOps” at this point, but I would never write an article like “What is DevOps?” because people with much deeper DevOps expertise than me have already described this more than adequately, as in The DevOps Handbook.

The challenge is discovering where you’ve internalized something that’s now obvious to you but non-obvious to enough other people to justify writing about it. I’ve developed a very simple technique to address this challenge: be self-aware of when I find myself explaining something over and over in different conversations. This is my signal that I’ve internalized something non-obvious and is the key input to potential writing topics.

Before committing to writing about such topics, I do one final test: research whether someone else has already written something good enough and I just haven’t found it yet. For instance, a couple of years ago, I found myself explaining to people that “Brooks’s Law” isn’t really a law but rather a set of dynamics that can be understood and, if you’re clever enough, defeated. So I started drafting an article called “defeating Brooks’s Law,” but after just a modicum of research, I found several articles on it, as well as an excellent broader talk on software process dynamics by Robert Smallshire, and an excellent textbook on the same topic by Raymond Madachy. So I never wrote that article but I’ve shared these resources several times since.

But sometimes I don’t find a satisfactory existing item and in that case, if I have the time and motivation, I commit to writing an article.

Scaling knowledge and expertise

When I do commit to writing my own article, I spend significant effort to research and write it. For instance, a few months ago I wrote an internal IBM article on the future of Watson, where I spent more than forty hours over the course of four weeks, given its importance to IBM.

It takes so long because—at least for me—the process of writing always reveals substantial points of ignorance and ambiguity of the topic. Often I’ll have to read other articles or books to address the knowledge gaps, and just figuring out how to say something concisely and precisely causes me to sharpen my understanding. So counterintuitively (though well-known to writers), the process of writing to teach causes you to learn a great deal.

Once I finish the article, I share it both on social media as well as on relevant Slack channels within IBM. But the article really comes in handy the next time someone asks me a question in the same basic zip code and rather than providing a rambling verbal response over the course of 15–60 minutes, I can immediately respond “read this, and let me know if anything doesn’t make sense, or if you want to discuss more relative to your situation.” It typically takes less than 30 seconds to share, and provides a much better knowledge representation, considering the care I took to write every word, as well as the fact that they can read it at their own pace, and have time to ponder certain non-obvious points.

As an example, as I gained expertise as a manager, I found myself first creating and then sharing a technique to help employees discover their intrinsic motivations so that they can align their work with their intrinsic motivations for greater happiness and success. After describing this technique in conversations with dozens of employees and other managers, I decided to write about it in my Intrinsic Motivation Exercise article, which I now share many times every year.

The meta purpose of this article

At the beginning of this article, I mentioned that I was writing this article for reasons you’d find clear by the end. I’m guessing it’s pretty obvious now.

As people I mentor are exposed to my body of writing, they often ask “how do you decide what to write about?” and now this article exists to answer that question! 🤘🏻

References

[1] The best simple primer I know on the topic of expertise is Anders Ericsson’s Peak: Secrets from the New Science of Expertise. Ericsson has many much more in-depth papers and books on this topic.