Deployment: After 450,000 Views

Deployment is a software term: taking something that works locally and putting it on a real server, where other people can actually use it. Once it's deployed, it's live. For writing, it's roughly the equivalent of hitting Publish.

01 / Exam | A Hundred Questions

Back in the day, you had to take an exam to register for a Bilibili account.

A hundred questions. Sixty to pass. Danmaku etiquette, anime trivia, community rules—everything was in there. Fail, and you had to wait through a cooldown period before you could try again. I sat through the whole exam just so I could post a video.

It was 2016. Short-form video was just taking off.

Looking back, the things I posted after passing the exam feel like they happened in another lifetime. I only remember one video, filmed during a visit to a panda base, that got over 400,000 views.

450,000. A panda's view count. I still remember that number today.

My best result after that was around 31,000. Most of the rest struggled to reach a thousand, usually hovering in the low hundreds. For a full-time UP creator, those numbers were nothing. But I spent countless nights chasing them. I never hit 450,000 again.

450,000 people watched a panda. No one cared who was holding the phone. It took me a while to understand that.

What hurt more was the way the tide went out. No notice, no explanation. Today the recommendation slot was yours; tomorrow it belonged to someone else. You watched the graph in your dashboard slowly flatten out from its peak, until it became a mound of dirt that would never move again.

The following year, short-form video exploded. I wasn't part of it.

That was the first time I learned: what happens after you go live is out of your hands.

02 / Farm | The Subsidy Years

Making videos was hard to sustain alone. Filming, editing, and publishing took far more out of me than writing. During 2015 and 2016, I was doing both, but gradually I gave up video and focused on writing.

Platforms were competing for creators. Toutiao had a program called “A Thousand Creators, Ten Thousand Yuan,” offering a monthly guarantee of 10,000 yuan to a thousand selected writers. WeChat public accounts were still riding a wave of growth. Penguin, Baijia, Dayu, Yidian News—new platforms kept appearing one after another. Looking back, Bilibili's entrance exam seemed almost ridiculous. The new platforms were practically begging you to join. There was barely any barrier to entry.

That monthly guarantee was for the big names. I was earning pocket change based on traffic: a few yuan for every 10,000 reads. For someone whose salary barely covered rent and food, that was still real money.

It took me a few days to figure out how things worked. In this game, whether the writing was any good came second. First, you had to get people to click. Even the finest wine goes unnoticed in a hidden alley. If no one opens the door, it can sit there forever.

I signed up for several accounts. The words in my headlines grew more abstract, more sensational: shocking, speechless, in tears, devastated. I gradually came to believe that whether someone clicked on a piece depended first on how hard those words hit. Questions had to be rhetorical. The most effective trick was to keep people hanging: “By the end, I was speechless.” There was nothing particularly speechless about it. I had made that part up.

Those headlines performed better than straightforward ones, and the money looked a little better too. But the rhythm was like feeding an appetite: make it more sensational today, then even more sensational tomorrow. Stop, and the numbers dropped. When the numbers dropped, so did that month's earnings.

I can say “I made it up” lightly now. But underneath those headlines were real people who had clicked through. Some had genuinely been fooled by “I was speechless” and asked in the comments, “What was so speechless about it?” My food money came from those clicks. That's a debt I still haven't quite dared to calculate.

Publishing was more exhausting than writing. Every platform had its own rules. Some flagged duplicate content, some sent pieces back for revisions, some approved them the same day, others made you wait. There were no tools to help. I did everything myself. I worked during the day, then spent the evenings formatting, adjusting, adapting, and publishing.

One night, I'd finished publishing a piece, and every platform had approved it. I opened each dashboard in turn. The numbers were unremarkable; the earnings column showed a few cents here and there. I didn't shut down the computer or write anything else. I just sat there refreshing the page, over and over. The numbers weren't going to move, but I kept refreshing anyway. Near dawn, I dozed for two hours, then headed out to catch the subway to work.

Eventually, I stopped publishing. I couldn't say exactly when it happened. Work gradually got busier, and writing slowly faded out of my life.

I had also written a few pieces that I genuinely wanted to write. I saved them in a folder and never published them.

03 / Own | A Place of My Own

I never deleted the old pieces. They're still on those platforms. Every now and then, I click through and look at them. I really was young back then. If a headline could get people to click, I could write just about anything. I still have the titles and publication dates from every platform saved in the cloud. Those were roads I'd traveled. Why delete them?

Now I write on my own website, a little site I built years ago using a project from GitHub. It's mine to manage. I may publish on other platforms again someday, but at least I get to decide on the headlines. AI now helps with the less interesting work of formatting and distribution, saving me more than half the effort. The time I get back is just enough to write something of my own.

You might say, “Even the finest wine goes unnoticed in a hidden alley. Aren't you still afraid of that?”

I am. For the first three days after publishing something I really want to write, I can't help checking the dashboard. I know there's nothing I can do by looking at it, but I look anyway.

Now that I'm no longer worried about making ends meet, the reasons for writing seem thinner. When I count them up, there's only one left: I want to write.

But I wouldn't say I've figured it all out. Between the person who stayed up until dawn for a few cents and the person I am now, telling himself he doesn't care about the numbers, there is probably nothing more than luck. I just happened to reach a point where I no longer had to worry about putting food on the table. If I'd still been in the same situation as back then, I'd probably still be writing sensational headlines.

There's one thing I can't verify: what would have happened if I'd published the pieces I really wanted to write back then? Not a single one of them ever went live.

That folder is still there. I don't know when those pieces will be ready for anyone else to see.

But that's only as of today. Tomorrow, if I'm afraid, I'll deal with it tomorrow.