I published forty-seven articles last year. Some were reported features that took weeks. Some were quick takes I wrote in an afternoon. A few were personal essays I'd been thinking about for months. I know exactly how many people read each one, because the analytics dashboard told me — and I wish I could unlearn every number.
Here's what I know: my most-read piece was a 400-word reaction to a celebrity breakup that I wrote in twenty minutes. My least-read piece was a 3,000-word investigation into labor conditions in the creator economy that took me six weeks of reporting. The internet doesn't care about effort. It cares about timing, about hooks, about whether the headline triggers the right dopamine response in the two seconds before someone scrolls past.
I used to think good writing would find its audience. That quality was its own distribution mechanism. I believed this the way people believe in karma — not because the evidence supports it, but because the alternative is too depressing to accept. The truth is that distribution is a completely separate skill from writing, and most of the best writers I know are terrible at it.
The analytics changed how I wrote. Not all at once — gradually, in ways I didn't notice until I reread my early pieces and felt a jolt of recognition. That person sounded different. Less optimized. More willing to take a weird structural risk or follow a tangent that didn't serve the headline. The later pieces were tighter, more shareable, more "effective" by every metric that matters to an editor or a platform. They were also less interesting.
I don't blame the platforms, exactly. They built systems that reward engagement, and engagement is a reasonable proxy for value — until it isn't. The problem is that the metrics become the definition of success, and once that happens, the writing starts serving the metrics rather than the reader.
There's a difference between those two things. A piece that serves the reader might challenge them, confuse them, slow them down. It might require a second read. It might not be shareable because the insight is too specific or too uncomfortable to repost without context. A piece that serves the metrics is designed to be consumed, reacted to, and shared — in that order, as quickly as possible.
I'm not quitting. I still think writing for the internet is one of the most democratic, exciting things a person can do. The barriers to entry are nonexistent. The potential audience is everyone. The form is still evolving. But I'm trying to rebuild my relationship with the work in a way that doesn't depend on the dashboard.
Forty-seven articles. Some good, some mediocre, some I'm proud of, some I'd rather forget. I don't know how many people read this one. I'm trying very hard not to check.
