Even TikTok Once Fought Hard for Every 10,000 New Users

Our teammate Piaf worked at ByteDance back in 2017, when Douyin was only beginning to take shape and TikTok’s DAU was as small as today’s early-stage apps—perhaps even smaller. She oversaw TikTok’s overall operations across multiple countries, which meant she witnessed, almost in full dimension, how TikTok grew from zero to tens of millions of daily active users. For most people in the internet industry, gaining such firsthand experience is incredibly rare. In truth, over the past decade, the number of products that started from nothing and grew to the scale of TikTok or Douyin can be counted on one hand.

Curiosity about that era has always stayed with me. I often asked Piaf what life was like back then—what happened day to day, what her workload looked like, when she got off work, whether the company provided dinner.

Her answers eventually formed a distant, blurry picture in my mind: it was the moment when ByteDance launched itself from the ramp it had built with Toutiao, leaping into a new and unprecedented growth curve. Countless users and endless streams of traffic all began from that turning point.

People often try to explain the success of Douyin and TikTok. Two narratives frequently appear. One believes ByteDance succeeded purely because of aggressive investment and a streak of luck—nothing particularly special, and anyone with enough money to burn could replicate it. The other argues the opposite: that TikTok’s triumph was the product of genius-level design, meticulous execution, and extreme rationality. In other words, follow the strategy, perfect the recommendation algorithm, and success is inevitable.

I’m hardly qualified to judge which is correct. But I often recall a conversation with Piaf as we were wrapping up for the day. I asked her what, back then, had triggered TikTok’s big growth. She thought for a moment and said, “Lip-syncing.”

I vaguely remembered that trend too. Around that time, many short-video apps—even ones like Xiaokaxiu—featured lip-sync content. For a while, lip-sync videos exploded in popularity worldwide, and people loved recording them.

So I asked her how many users that feature brought in. “Around ten thousand new users a day,” she said.

I was stunned and asked again to confirm: ten thousand, not one million or five hundred thousand? As someone responsible for TikTok across multiple countries, she believed those ten thousand new users a day were incredibly important. Eight years later, she still remembers that number clearly.

I know logically that gaining ten thousand new daily users is never easy for any product. But this was TikTok—a platform that now serves a billion users every single day. And yet, back then, that ten-thousand-user boost mattered deeply.

In our own work, we’ve seen single-day organic surges surpass ten thousand more than once. And on today’s AI frontier, growth at multiple orders of magnitude frequently becomes headlines. We’ve become used to rocket-ship narratives—instant fame, instant glory, and instant collapse.

Like a farmer obsessing over how an emperor uses a gold hoe, I kept asking what her team actually did during those early days. Why did TikTok ultimately take off so dramatically? What did “0 to 1” feel like at ByteDance? She thought for a long time, then later that night messaged me after revisiting her old chat logs:

“Back then, we were so lost and exhausted. The path we now call ‘the great lift-off’ felt painfully messy and chaotic at every single step. Time and success have simply polished over the suffering. In 2017, when I joined TikTok, there was no talk of globalization or grand narratives. The entire overseas operation grew out of barren ground. Two PMs and a group of fresh graduates doing ops. Even the English translations in the backend were often embarrassingly wrong. Our earliest form of ‘globalization’ was literally proofreading English in the product.”

So how did you do cold starts back then? I asked.

She laughed and rubbed her forehead. “At that time, DAU in each overseas country wasn’t even ten thousand. The cold start was so cold that our biggest daily task was scraping huge volumes of content from various video platforms, removing watermarks, deleting subtitles, and re-uploading them. Crawler content made up the bulk of everything. Meanwhile, we were signing local creators, running shadow accounts that reposted top Douyin and Korean TikTok videos, and carefully testing how these influenced local ecosystems.”

What did your day-to-day work look like? I pressed.

“We pulled creators, ran challenges, pushed for submissions; we manually reviewed videos one by one to maintain a consistent style; and we watched competitors closely—who climbed the charts today, who poached creators tomorrow.”

Piaf continued: “Inside the team, it was chaos. Endless work. Explosive emotions. Arguments, breakdowns, insomnia—completely normal.”

In 2018, Piaf was preparing to travel to Vietnam for research when her colleague Nicholas told her he was considering resignation. They had fought side-by-side for a long time. She shared the chat logs with me—she had offered him heartfelt encouragement.

“I never again said anything that sentimental to any team,” she confessed.

Not long after, TikTok Vietnam’s DAU hit 300,000, then quickly passed one million. Vietnam became another “successfully launched” market for her team. I applauded her wholeheartedly, and she kept sharing:

“We continued reviewing content day and night, shaping the style, syncing top Douyin creators via shadow accounts. We realized something big—Chinese pop culture influenced Vietnam far more deeply than any other country.”

So the team made what she called a “wild” decision: they flew top Douyin creators directly to Vietnam to compete with local creators offline. It was TikTok’s first large-scale offline event in Vietnam and the first official cross-border creator collaboration.

Among those creators was Zhang Xinyao, one of the biggest stars at the time.

No one knew this event would later become the starting point for TikTok’s global creator collaboration model.

The event was a huge success. Online content exploded.

But Piaf and Nicholas didn’t get a single proper meal that day. By the time they finally sat at a roadside stall late at night, their phones buzzed.

Kwai had surged to the top of the charts.

The next battle had already begun.

In April, TikTok released the Duet feature. Combined with challenges, submission rates skyrocketed. Meanwhile, Kuaishou’s overseas version went on a poaching frenzy—targeting every major creator and even Nicholas.

“I knew it was tough for him,” she said. “But I also knew his loyalty to the team meant more to him than any offer.”

In May, Vietnamese competitor Muvik was taken down. One less enemy—yet their workload didn’t ease.

In July, the Chinese drama “Story of Yanxi Palace” went viral in both China and Vietnam. The team launched cross-regional campaigns, and summer content creation exploded. At the same time, Vigo, ByteDance’s other product, withdrew from Southeast Asia and merged into TikTok’s operations.

The Vietnam team grew larger. But the workload never shrank.

They still had late-night breakdowns. They still questioned everything at dawn.

But amid the exhaustion and chaos, Vietnam’s content ecosystem slowly took shape.

Retention rose from 35% to 55%, the highest among all TikTok overseas markets. And after Indonesia’s shutdown, Vietnam became the new growth engine.

But back then, none of them looked up long enough to notice these milestones.

They simply kept their heads down and pushed forward.

At this point, I noticed she mentioned Indonesia’s shutdown—a hardship I had long been curious about. So she continued:

From June to July 2018, TikTok was banned in Indonesia for two full months.

And Indonesia was the world’s fastest-growing and largest market at the time.

Six million DAU vanished overnight—back to zero.

“I was responsible for Indonesia, Vietnam, and Thailand. There was no escaping that blow.”

They checked processes, people, mechanisms—suspecting even deliberate sabotage—but found no real cause. By the time they discovered the truth, the ban had already taken effect.

And back then, they didn’t even have a dedicated government relations team.

“There was no warning. No buffer. No negotiation.”

Piaf remembers her reaction vividly. “It wasn’t emotional. It was the cold realization that everything you built can be paused by the world without giving you a single second to respond.”

She barely slept during those two months.

Looking back, she now believes the setback saved TikTok.

“The ban forced us to understand—GR and PR aren’t optional overseas. They’re survival. TikTok’s global GR team began because of Indonesia. If we hadn’t paid that price then, far bigger disasters would have come later.”

By now, night had fully settled outside the office. After a long pause, she added:

“Nicholas talked to me almost every day during that period. He told me, ‘No matter where we are, we’re walking a path no one has walked before.’ And I said, ‘Yes… except some of the holes we fall into cost millions of users.’”

People now remember TikTok’s soaring retention in Vietnam, its rebound in Indonesia, its rapid rise in Southeast Asia.

But very few mention those two months of “disappearance.”

And Piaf rarely speaks about it either—because it remains the most powerless chapter of her career.

We were getting hungry by then and ordered KFC. I still remember they had just launched a new item—the Crispy Golden Chicken Wing. We loved it, and then it was discontinued the next day. It has never returned. A trivial detail, but as I write this, I miss it dearly.

Everyone knows what happened afterward. Once the massive boulder of the recommendation algorithm had been pushed to the mountaintop, its momentum rolling downhill was unstoppable.

But that boulder didn’t climb on its own. It was pushed up—by ten-thousand-user increments, by global hustle, by countless manual video adjustments, settings, selections. More than that, it was built on sleepless nights, emotional crashes, chaotic rhythms, and the fragile space between “I can’t hold on” and “just one more push.”

All of it dissolved into the halo of a single success story.

We often beautify outcomes or simplify narratives. I used to be cautious about retrospectives because of this. But my perspective has changed: beautified or simplified, it doesn’t matter. No one can restore the full truth anyway. If a story allows us to grasp even one meaningful insight—one spark that resonates—then it has fulfilled its purpose.

For me, hearing these stories from Piaf’s perspective made me genuinely admire ByteDance. Not because of its money or its hit products, but because in the very beginning there truly was a team full of energy, grit, and willingness to dive head-first into the unknown. That team included Piaf, Juanjuan, and many others who are now senior leaders at the company.

Today, we stand in the roaring tide of AI—discussing compute, models, emergence. Everything seems infinitely accelerable. And maybe that’s true. But failure also accelerates. The path we walk now is different from before, yet the lack of shortcuts remains unchanged.

What we need is patience, the right teammates, and the determination to keep our heads down and push forward.

Wind first rises as a gentle breeze.

And today, it is still only the beginning.