This Woman Is Manually Farming Views on a Phone Rig So Someone Can Pretend They’re Famous Is Just Depressing And Sad

There’s a video making the rounds of a woman in a fluffy white cardigan sitting at a desk in what looks like a normal office, methodically tapping and swiping across a custom rig of more than a dozen smartphones locked into a mechanical frame. The phones are all running video feeds, mostly short-form content that looks like TikTok or Reels. She’s not doom-scrolling for fun. She’s working. The caption on the post that pushed it: “Getting those social media views up!” The crying emoji is doing a lot of work.

This is a click farm, or view farm, or engagement farm—whatever branding the people running it prefer this week. It’s not a sci-fi botnet of pure software ghosts. It’s a person, on the clock, generating the illusion of an audience so someone else can look more popular than they actually are. The replies under the video called it depressing. One said a robot should be able to do all of it. Both takes are correct.

The setup is familiar if you’ve ever gone down the rabbit hole on this stuff. Farms like this have operated for years in places with cheap labor—Vietnam, parts of China, Thailand, India—where a room full of phones and a few people can be more cost-effective than pure bots that get flagged immediately. Real devices, real SIM cards or residential proxies, real human fingers (or human-controlled software that still needs someone watching the screens) produce traffic that looks less suspicious to platforms that have gotten better at spotting pure automation. One person can work a bank of phones. Scale that across shifts and you get volume.

The business model is simple and ugly. Influencers, brands, musicians, e-commerce sellers, and the occasional political operation buy packages of views, likes, follows, and comments. According to recent pricing data pulled from services that openly sell this stuff, views are dirt cheap: roughly $1.40 per thousand on TikTok and up to about $6.70 per thousand on YouTube. Likes run $10–25 per thousand across major platforms. Followers are in a similar range on most apps. Comments cost more because they’re harder to fake convincingly. The farms buy the labor and infrastructure cheaply, sell the output at a markup, and do enough volume that the margins work even when individual orders are tiny. Workers often get paid fractions of a cent per interaction. The person in the white cardigan is not getting rich off this.

Clients buy it because the platforms’ algorithms still heavily reward engagement signals. A video that suddenly has thousands of views and likes gets pushed to more real people. Social proof helps land brand deals, sponsorships, and the perception that you’re “blowing up.” Some of it is pure vanity. Some of it is calculated: inflate the numbers enough to get into the real recommendation engine, then try to convert a portion of the resulting organic traffic. Platforms ban the practice in their terms of service and periodically purge accounts and adjust detection, but the demand doesn’t disappear. When real engagement is hard to get and the rewards for appearing popular are high, a gray market fills the gap.

The video feels grim because it strips away the abstraction. It’s not an anonymous botnet. It’s a person doing repetitive physical work so that somewhere else an account can claim an audience that doesn’t exist. The phones are arranged so one worker can reach them efficiently. There’s a keyboard, cables, people walking around in the background like it’s any other office. The content on the screens is the kind of stuff the algorithm already loves. The whole thing is optimized for throughput.

This industry has been around long enough that the romantic version, lone hackers in basements, is mostly dead. What remains is industrial: racks of devices, low-wage labor, software that helps one person control more phones, and clients who treat fake engagement as a marketing line item. Platforms keep playing whack-a-mole. Detection improves. Farms adapt by using more real devices, more human-looking behavior, and more distributed setups. The numbers stay cheap enough that people keep buying.

The crying emoji on the original post is honest. Watching someone manually generate fake popularity is a pretty bleak window into how the attention economy actually works when the incentives point toward manufacturing the numbers instead of earning them. The robot comment under the video is probably right that more of this will get automated. Until then, someone’s still sitting there swiping.