After 10 Years in Operations, I Hit 800K Monthly Revenue—But Also Saw 50+ Plugins Delisted, 19 Accounts Banned, and 250K in Penalties

Startup Newbie Protection Period - Part 1 When money comes too easy, it’s tempting to credit all the era’s dividends to your own ability.
Hi everyone, I’m Fantexi.
Today I want to share my entrepreneurial story.

I spent nearly a decade as an employee before quitting to start my own business, eventually running two companies.
I built a mini-program with nearly 20 million cumulative users, and across all products, my company’s monthly revenue peaked at over 800,000 yuan.
I also made plenty of mistakes: over 50 Chrome extensions were delisted overnight, 19 Chrome developer accounts were banned in succession, and I was sued over an old, unmaintained project—resulting in a court-ordered penalty of about 250,000 yuan.
At one point, I also had several WeChat service accounts with around 30,000 followers each, and one with 300,000 followers.
But what I regret most is not recognizing the value of self-media and personal branding sooner, nor properly nurturing those user relationships.

Now, I’ve decided to pick that back up.
Starting with my own journey, I’ll honestly document how I made money—and how I gradually squandered those advantages.
If you’re building products, starting a business, or going global, I hope these experiences help you avoid pitfalls and offer some inspiration.
This first part explains why I left my nearly decade-long Linux system admin career and built a company with monthly revenue exceeding 800,000 yuan.

After Nearly 10 Years as a Linux Sysadmin, I Finally Saw My Ceiling

Before 2018, I switched jobs a few times, spending almost a decade in Linux system administration.
I then joined a locally based A-share listed company with over 500 employees, leading a Linux operations project.
The income wasn’t bad—about 17,000–18,000 yuan per month, plus year-end bonuses, totaling over 200,000 yuan annually.
The job was stable.
But I could clearly see my ceiling.
Bluntly put, I was just another cog in the machine.
Without me, the project would keep running no matter what.
No matter how well I did, I was just tightening a screw a little tighter.

The company still relied heavily on traditional physical data centers, and project migrations often happened at 3 or 4 AM. My health couldn’t take it anymore, and my technical skills hadn’t improved noticeably for years.
I felt I wasn’t accumulating experience—I was depleting myself.
Around that time, a friend told me about a company with fewer than 20 people that had made a casual game with great feedback.
When I went for the interview, the room was just big enough for two people, with cardboard boxes stacked nearby.
No big-company prestige.
No fancy office.
It even felt a bit like an amateur production.
But one of their games had held the #1 spot on China’s App Store free chart for over 60 consecutive days.

I left the listed company and joined them.
On my last day, I ran into my former manager in the elevator. He was puzzled: why leave a solid listed company for a small startup?
He didn’t know—I wasn’t just changing jobs.
I wanted to learn how to operate consumer products, drive user growth, make product selections, and understand how founders think about users, assess needs, and make trade-offs.
I wasn’t there for a promotion.
I wasn’t there for stability.
I was there to learn the craft.

The scariest thing about stability isn’t that it keeps you from making money—it’s that it slowly conditions you to accept that this is all your life will ever be.

Users Keep Paying, But the System Could Lose the Day’s Data Any Moment

After joining the company, I saw a Redis server.
If you’re not technical, you can think of it simply as: the product only saves a full backup once a day. If the server crashes midway, all of the day’s user transactions and data could vanish.
My first reaction was: that’s incredibly unprofessional.
From an ops perspective, this architecture was practically running naked.

What struck me even more, though, was that despite this terrible architecture, users were paying into it every single day.
Users don’t care what tech stack you use.
They don’t care if your code is elegant.
They only care whether your product solves their problem.

That was the first time I truly realized:
A poorly built system that people pay for is closer to the answer than a perfect system nobody uses.

Back then, I only learned the first half: tech isn’t everything.
Years later, when my extensions were delisted by Google for security and compliance issues, I learned the second half:
Tech might not decide whether a product can launch—but security and compliance can decide how it dies.

I Couldn’t Write Mini-Programs or Handle Promotion—So I Found Users One by One on Baidu Zhidao

In mid-2018, I spotted the opportunity in the WeChat mini-program ecosystem.
The problem was, I didn’t know how to build mini-programs.
I bought a development book and secretly coded after work, just enough to understand the basics.
Should I wait until I learned everything from scratch before starting?
If I waited until I knew it all, the wind might have already died down.

I found a former colleague to partner with. I handled server setup, product selection, and promotion; he handled the coding.
We built several mini-programs, including ones for event registration, iOS rankings monitoring, couple avatars, and watermark removal.
The couple avatar one initially performed just okay on WeChat, but I thought it had two key traits:

• Couple avatars come in left-right pairs, requiring two people to find and match together.
• Once found, the natural next step was for users to share the other image with their partner—making it inherently viral.

The watermark removal tool actually got better feedback—but I wouldn’t recommend that path now. Years later, we paid the price for it.

After building the products, I went looking for users.
I didn’t know growth hacking or brand marketing.
I just knew users were asking questions on Baidu Zhidao and Baidu Jingyan.
So I went there to reply.
I searched question by question.
I answered user by user.

Eventually, daily revenue grew from a few hundred yuan to around 3,000 yuan.
I remember when the first user paid 300 yuan—I was thrilled.
Not because 300 yuan was a lot.
But because it was the first time I confirmed that people were willing to pay real money for something I built.
These users weren’t just from China—they came from Japan and Malaysia too.

Later, the boss found out I was doing this side project, and it was time for me to leave.
After I quit to start my own business, I rented a shared desk.
Surprisingly, I felt no anxiety at all during that time. I’d work in the morning, and sometimes go to the movies alone in the afternoon—often the only one there, like a private screening.
With money coming in every day, I felt for the first time that maybe I could truly make a living from my own products.

The Product Barely Took Off on WeChat—But Hit Nearly 20 Million Users on QQ

The real explosion happened in the QQ mini-program ecosystem.
When QQ’s mini-program platform first launched, I secured one of the initial spots.
The first thing I did was migrate the couple avatar mini-program over.
The logic was simple: QQ’s users are younger, so couple avatars might resonate even better with them.

The product launched and within two or three days hit the top three on the new releases chart. It later made the overall top ten and consistently ranked high in its niche.
Combined with the surge in internet traffic during the pandemic, this mini-program reached nearly 20 million cumulative users, with daily ad revenue exceeding 5,000 yuan at its peak.

That was just one of our products.
We built a “Guess the Celebrity” mini-program too. In the first week, it had under 1,000 daily active users; by the second week, it had jumped to 100,000.

Later, when QQ’s avatar customization platform opened to third-party developers, I was one of the first to experiment.
Initially, producing one avatar cost over 700 yuan, and we needed at least 20 a month. For several months, the project generated revenue but barely made a profit.
As the platform matured and more professional designers joined, I gradually figured out product selection and talent recruitment.
Monthly revenue soon jumped to 200,000–300,000 yuan, recouping all previous months’ investment in a single month.

I then entered the <> map ecosystem. This project needed only two to three people and achieved monthly revenue of 100,000 to over 200,000 yuan.
When one product made money, I used that revenue to fund the next.
When the next one made money, I hired more staff.
A fresh graduate joined, and by their second week, they were generating revenue exceeding their monthly salary.

Eventually, across all products, the company’s monthly revenue peaked at over 800,000 yuan.
At the time, I really believed:
I was born to be an entrepreneur.

Looking back, what I actually knew then was: product selection, seizing windows, finding users, and pushing a workable product forward.
What I didn’t know was: management, compliance, long-term operations, and turning a temporary project into a transferable asset.

The company expanded too fast, and my management skills didn’t keep up. I provided housing and broadband for employees, but failed to establish clear goals, collaborative workflows, or growth mechanisms.
I could push products up—but I still didn’t know how to guide them through a full lifecycle.

I Thought I Had 5 Products—But Really, I Only Tapped Into One Type of Advantage

At that time, I had several projects: couple avatars, watermark removal, celebrity guessing, QQ avatars, game maps, and more.
I thought, if one project failed, others could carry the load.
That’s called risk diversification.

Later, I realized these projects served similar young users and relied on similar platform traffic.
They looked like multiple products, but their roots were all in the same soil.
As long as the underlying dependencies don’t change, five products might just represent one risk.

Back then, I didn’t realize how much that would cost me.
Growth stalled. Over 50 Chrome extensions were delisted overnight. The first 19 developer accounts were banned. By 2025, an old project I hadn’t maintained for years resulted in a compensation claim of about 250,000 yuan.

Only then did I understand: monthly revenue of 800,000+ yuan didn’t mean I’d learned how to run a business.
It was just a newbie protection period that entrepreneurship had granted me.

Next Up: How Did My Startup Newbie Protection Period End?

After hitting 800,000+ yuan in monthly revenue, the real test was just beginning.

• After policy changes, why didn’t the business die immediately—and why did I realize the danger even later?
• Why were over 50 Chrome extensions delisted overnight?
• After the first 19 developer accounts were banned, why did I still refuse to stop?
• Why did a project dormant for years suddenly come back in 2025 with a million-yuan compensation claim?
• How did I shift from chasing traffic to refining SiteData with real users?

These experiences taught me a key distinction:
Knowing how to seize an opportunity and knowing how to build a company are two entirely different skills.

If you’re also building independent products, taking tools global, or standing at the crossroads from tech to business, feel free to follow me.
I’ll write about the successes—and the lessons paid for in tuition.

A Gift for Readers:
If you’re interested in SiteData, I’ve prepared 100 lifetime 30% discount codes (redemption code: fantexi). This offer is placed only at the end of this article—just for those who’ve read all the way here.