For about eight months I spent three evenings a week talking to almost nobody. I had a microphone, a webcam pointed at a bookshelf I had tidied specifically for the camera, and a small number in the corner of my screen that told me how many people were currently watching. Most nights that number was zero. Sometimes it was one, and the one was often a bot that scraped channel titles.
I want to describe what that does to a person, because I think it is a more common experience than men usually admit, and because the shape of it shows up in a lot of work that has nothing to do with streaming.
The specific humiliation of a live counter
Writing into a void is an old experience. Painters, novelists and garage bands have all done work that nobody asked for and nobody saw. What is new is the counter.
When you write a novel nobody reads, you at least get to finish it before you find out. Live streaming removes that mercy. The audience number sits on your screen the entire time, updating every few seconds, while you are mid-sentence, trying to sound like a person who is enjoying himself. You are performing enthusiasm and reading your own failure at the same moment, in the same field of vision.
I got good at a particular kind of split attention. Half of me would be talking about whatever I was doing, and the other half would be running a quiet commentary: still zero, still zero, someone joined, they left, that was four seconds, what did I say in those four seconds, was it the thing about the weather.
That second voice is the problem. Not the empty room. The room is fine. The room is just a room.
Why the number gets into your chest
There is a reason a viewer count lands harder than most feedback. It is public, it is continuous, and it appears to be a measurement of you.
None of those three things are actually true, but they feel true. Public: anyone who clicks on the channel sees the same number you do, which means your lack of an audience is itself broadcast. Continuous: it never stops reporting, so there is no protected space in which to just do the work badly for a while. And the last one is the real trap. A viewer count is not a measure of quality, or effort, or whether the thing you made was worth making. It is mostly a measure of distribution.
Distribution is a system, not a character trait. But when the only signal you receive is a number, the number becomes the whole story, and men in particular are practiced at converting an ambiguous signal into a verdict about our own worth. I know how to work hard. I did not know how to work hard at something that reported back, in real time, that nobody cared.
It also compounds in a way that felt personal but is purely mechanical. A low viewer count keeps a channel low in the directory listings, which keeps the count low, and the loop closes on itself. Rationally I understood this the entire time. Emotionally, at eleven at night with an empty room and my voice getting thinner, I did not experience it as a ranking problem. I experienced it as a referendum on whether I was interesting, and I lost it three nights a week.
By month four I noticed I had started making decisions to move the number rather than because I wanted to. I changed what I streamed. I streamed at hours that were bad for my sleep because that was when the category was less crowded. I stopped doing the slow, quiet, technical parts of the hobby that I actually liked, because they made the graph flatter. That is the real cost. Not the sadness. The slow substitution of your taste for someone else’s algorithm.
The industry that grew around the anxiety
Anyone who has felt this has also seen the offers. Search any version of “how do I get more viewers” and you land in a market built on top of exactly this feeling.
Some of it is honest. Coaching, thumbnail help, editors who cut your streams into clips, schedule advice. Some of it is automation. There is a whole category of paid services, and this Twitch growth tool is one of many that operate in the same space, that will put viewers, followers or replay views on a channel for a monthly fee. They exist because the discovery systems on the big platforms reward channels that already look busy, which creates a genuine cold-start problem and a genuine market for solving it artificially.
It is worth being blunt about the tradeoff: inflating the numbers on a channel runs against the terms of service on every major platform, and anyone considering it should understand they are putting the account at risk. That is the practical objection.
The personal objection is different and, for me, larger. If the number is what hurts, then buying the number treats the symptom. You still walk into the room every night hoping to feel like the work mattered, and now the evidence has been tampered with by you, personally, which is a strange thing to try to believe.
I am not interested in moralising about it. Men do worse things than pay for a metric. But I know myself well enough to know that the version of me who bought viewers would have needed a bigger number a month later.
What actually helped
Three things, none of them clever.
I moved the counter. I hid the viewer count from my own screen and only checked it after the stream ended. It was the single biggest change I made. Two hours of talking without live scoreboard access turned out to be a completely different and much better activity. The recordings from those months are noticeably better, because I was doing the thing instead of monitoring the thing.
I picked a metric I could actually control. Viewers are downstream of a hundred factors I do not touch. Whether I showed up on the days I said I would show up is entirely mine. So that became the number I tracked: sessions completed, not people present. It is a smaller, less flattering measure, and it is the only one that ever moved because of something I did.
I named who it was for. Early on I would have said “anyone”, which sounds generous and is actually a way of avoiding commitment. When I got specific about the handful of people who might genuinely want this particular thing, the empty room stopped feeling like a rejection by the general public and started feeling like what it was: a niche, in its early days, with a slow on-ramp.
The part I did not expect
Somewhere in month six I had a stream with four people in it, and one of them said something in chat that made the other three laugh, and I remember thinking clearly that this was already the thing. Not a step toward the thing. The thing.
Four people who chose to spend an evening in a room with you is not a consolation prize for not having four thousand. It is a specific and fairly rare kind of good. The counter cannot tell the difference between four people who are present and four hundred with a muted tab in a background window.
I have come to think most audience metrics are like that. They are cheap approximations of something real, and if you stare at them long enough you begin optimising the approximation instead of the real thing. That is true for streamers and it is also true for anyone whose work now comes with a public dashboard, which is most of us.
Where I landed
I stream less than I did, and I enjoy it more. The channel is still small, and I have stopped organising my week around the possibility that it might not be. What those eight months gave me was not an audience but an uncomfortable education in how fast I will trade my own judgement for a visible score.
If you are talking into an empty room right now, I would not tell you the audience is coming, because I do not know that and neither does anyone selling you anything. What I would say is that the room is not a verdict. It is a room. The counter is not measuring your worth. It is measuring distribution, badly.
Cover it up. Do the two hours. Judge yourself on whether you showed up and whether the work was any good, which are the two questions you are actually qualified to answer.
