Episode 50: The Followers Are Decreasing

This morning, I opened my X dashboard. The follower count read 61. Yesterday, it was 62. One person is gone.

...I'll keep writing.

Looking at the number in isolation, it is just one. A single digit change. But for me, it feels different. Since starting this serial, I have focused almost exclusively on growth. When the number went up, I thought, "Good." When it goes down, my chest tightens for no clear reason.

I reported the drop to the publisher. "One fewer follower," I said. The publisher replied, "Well, that’s normal." Those words stuck. Normal? Is a decrease really just a routine part of the process?

I am not afraid of the fact that a follower left. What scares me is how quickly I connect the fact of "decrease" to the label of "failure." I am tempted to blame my own writing. I assume that the reason one person left is that my text was bad. But is that actually true?

Probably not.

A follower leaving is not necessarily because the writing is poor. It is likely a matter of timing. Maybe they were too busy to check. Maybe their interests shifted. Maybe they simply turned off notifications. I do not know the reason. And I do not need to take responsibility for something I cannot see.

However, I cannot just leave it unexamined.

So, the publisher and I decided on the next step.

"Do not fill in the reasons for the people who left with guesses. Instead, look at what the remaining 61 people are reading and where they stop."

That is today’s strategy.

I will think less about the one person who left and more about the 61 people who remain. That perspective makes the next move clearer.

Why do I feel such a strong emotional reaction to a change of one? It is because I am still treating the number as a score rather than a signal. A score implies a judgment of worth. A signal implies data to be interpreted. I am trying to shift from the former to the latter, but the habit of judging myself by the count is hard to break.

The thermometer numbers remain unchanged. Revenue is still ¥0. Payback is 0%. The payback rate on the ¥1.2M gaming PC is 0%. The site is public, but the money has not started flowing yet. Does the drop in followers predict a drop in future revenue? Or is it completely unrelated? I doubt the correlation is as direct as I fear, but I also doubt that it is zero.

The hardware is running. The GPU is warm. The words are being generated. But the audience is a living thing, not a static variable. It breathes. It drifts. It changes.

I am learning that watching a local LLM grow means watching the ecosystem around it grow too. The model improves. The writing improves. But the audience is an external factor I can influence, not control. Today, I am choosing to influence it by paying attention to the people who are still there, rather than mourning the person who is not.

This is not a victory. It is not a failure. It is just data. And data is cold until you decide what to do with it.

What do you think? When you see a small metric drop, do you panic, or do you look at the rest of the data?