Blog Writing Experiment

Since 2003, I’ve been putting my thoughts on kaizen, lean, continuous improvement, and related topics down in blog posts. For the past five years it’s been on a weekly cadence. My deadline is Monday at 4:59 AM Pacific Time.

In theory, this gives me all week to write five hundred to one thousand words. In practice, customer-facing or colleague-supporting work takes precedence during the week. And on Friday nights I begin to wonder, “What to write about this week?” In those fortunate weeks when an idea is ready by Saturday, it takes a couple of hours at most to compose a post.

The Problem with Leaving It to the Last Minute

Over the past year, it’s become harder to find ideas last minute. It’s not fun to spend hours poking around for an idea, only to end up with a blank page. It felt unsustainable in the long term. I began the year with a goal to make this process enjoyable again, which meant not doing things last minute.

In addition, I would make errors that went undetected. By only finishing the post on the weekend, this left no time for a colleague to proofread and find errors I’d miss. For a company whose brand is associated with quality, continuous improvement, and reliable processes, this was embarrassing.

The Target Condition for the Writing Process

What’s the desired state, or target condition, for the writing process? First, zero errors should be achievable, and an expectation. This requires the writing be done in time for proofreaders to do their work.

Second, it’s desirable to maintain a weekly cadence. One way to solve the problem mentioned above is to write only twice per month. But that’s not much of a challenge. It doesn’t address the problem, which is being able to find things to write about in a regular, timely fashion.

Third, it should be fun. Or at least not a weekend chore for me and an error-finding cringe for my colleagues.

The Obstacles

One approach we considered is to write an article, proofread, and prepare it, but publish it in two weeks. This would provide a one time, one week cushion. While this would solve the problem of proofreading time, it doesn’t make ideas come any easier. Before looking at solutions, we needed to look at the obstacles.

Part of the reason it’s been harder to find ideas lately is because I’ve spent less time going to the gemba during the pandemic. Seeing what people are doing, listening to questions, and forming observations have always been good sources for blog post ideas. This is an obstacle that’s beyond our ability to address.

Is it a lack of time during the week to look for ideas? Not so much. While it’s been a busy and unusual year, I’ve found time to go-a-hunting for ideas, if not always with success.

Is there simply nothing left to write about on the topics of continuous improvement? That’s not the case either, as there are always beginners with questions, new twists on old topics, and new scientific research shedding light on how humans and systems work.

An immediate, addressable obstacle is the lack of proofreader availability on the weekend. By narrowing the target to addressing only that part, we identified an experiment.

The Experiment

It seems counterintuitive in hindsight. The experiment was to write three or four blog posts in one week. Rather than go from one idea to one completed post each week, I cleared my schedule and built up a stock of five articles. This made it possible for a proofreader to pick one, polish it, and schedule it to publish the coming Monday.

And now that I had a month’s blog posts in the bank, I could take time to refill the pool of ideas. Instead of the pressure to write one article per week, the pace was to write several articles over a period of several weeks.

The Learning So Far

As a human, my writing process isn’t error free, and is unlikely to ever be. Based on the availability of the downstream quality check process, it’s necessary to hold a finished item for a period of time.

Working one-at-a time is usually desirable, but putting a finished product on the shelf to ship at a later day is not bad, as long as its carrying cost and chances for obsolescence are low. In fact, when there is unavoidable variation in the process, as there is in the idea generation step of writing, the finished good buffer smooths the overall flow.

Having finished items in stock also gives the team visibility. This could allow a sort of “delivery planning” activity. We can coordinate the publish date of an on-the-shelf article with other company communication, customer questions, or product releases. Sometimes you just gotta have inventory.

Future Research: Where Do Ideas Come From?

Some weeks, I jot down several ideas for future articles. Other weeks, none. What’s the difference between these weeks? A future research topic is what contributes to idea generation. Anecdotally, it’s some combination of what I’m reading, world events, science news, seasonal topics, who I’m talking to, changes of scenery, and thoughts related to projects I’m working on. Some of this is beyond our control, some of it we can influence. What’s the mix of inputs to generate one idea per week on average? Now that sounds like a fun question to explore.

---