Chapter 28: I Am the State

Entertainment Savior A commoner from eastern Zhejiang 4538 words 2026-03-25 00:17:55

What Gu Cheng referred to as "creating a federal deposit insurance company for the internet industry" was, in essence, to establish an intermediary platform for online advertising akin to AdSense.

The operational principle of this platform is roughly as follows:

Before AdSense emerged, most websites worldwide charged advertising fees based on third-party traffic statistics—such as those provided by Alexa—to assess the value of a webpage’s ad space. The daily fee for a particular ad slot displaying a specific advertisement was fixed. The number of clicks did not directly affect the website owner’s income. (If it did, the situation would be worse because fabricated clicks could easily occur, resulting in heavy losses for advertisers.)

However, the accuracy of third-party traffic evaluation systems, including Alexa, was utterly destroyed by fabricated daily active IP counts before the industry downturn. Consequently, advertisers worldwide stopped trusting any website owners, doubting the effectiveness of internet ads and considering the inflated premiums unrealistic. Many preferred returning to traditional media like TV, radio, and newspapers for advertising.

As a result, global internet ad valuations plummeted, ushering in a harsh winter.

Then came the AdSense platform, which transformed the model: advertisers no longer dealt directly with website owners. Instead, they submitted their advertising demands to AdSense. Similarly, website owners with vacant ad spaces reported their available resources to AdSense, which then matched supply with demand.

AdSense employed a transparent, publicly credible matching algorithm, gaining public trust and oversight. Fees were charged based on effective clicks—each IP address counted for at most one valid click per 24 hours. The platform’s algorithm also judged whether a click was genuine or maliciously fabricated.

This intermediary system dramatically improved advertising efficiency.

Initially, when website owners and advertisers negotiated directly, each site’s ads were fixed and often failed to maximize display efficiency.

For example, the Los Angeles Lakers’ official website might only attract advertisers like Nike and Adidas selling shoes. If visitors to that site were unlikely to buy shoes, the ads were ineffective.

Once entrusted to AdSense, supported by big data and search engine algorithms, the platform could analyze that basketball and jersey ads had the highest click rates on that site.

Google would then search among multiple advertisers with basketball ads, some bidding five cents per click, others twenty. Naturally, Google would prioritize pushing the twenty-cent ads.

However, AdSense’s algorithm did not simply rank ads by bid price per click; it also maximized the combined product of unit price and expected click volume.

For instance, besides basketball gear, products like Yunnan Baiyao, a medicinal ointment for bruises and sprains, also had high demand among sports site users. Being a consumable, it sold more frequently than basketball equipment.

If a Yunnan Baiyao ad offered five cents per click but had a click-through probability more than four times higher than basketball ads, AdSense would prioritize Yunnan Baiyao despite its lower bid.

Thus, this intermediary benefited both advertisers and website owners. Neither party wished to eliminate the middleman; on the contrary, they welcomed one that ensured mutual benefit.

This brief overview only scratches the surface of the ad algorithm’s complexity; writing the full essence would require hundreds of thousands of words.

Historically, Google officially launched AdSense around mid-2003, monetizing it. Google’s rise to a multi-hundred-billion-dollar giant relied heavily on this cash cow: for every valid internet ad click worldwide, Google collected one to two cents in intermediary fees, totaling billions annually.

Gu Cheng didn’t need to memorize all this history. He only needed to know that Google had yet to perfect this system fully, and if he and Mr. Li wanted to build it, the core technology was already mature.

Three key elements were essential to building an AdSense-like platform:

First, a robust and continually improving algorithm—this was crucial for credibility, since click fraudsters, hackers, and ranking manipulators constantly evolved their techniques. The platform had to upgrade its defenses around the clock to maintain an edge.

Just as Taobao and certain Chinese content sites fought ranking fraud with strong technical teams refining detection algorithms.

In 2006, a fraudster on Taobao might simply register a new account and purchase from the targeted store to fake transactions.

By 2016, the process became far more complex: newly registered accounts had to remain idle for 15 days, conduct multiple balance transfers with other users in Alipay, simulate conversations with store clerks, browse several similar product links, and limit monthly purchases to avoid detection.

This decade-long battle between fraudsters and platform defenses drove continuous technical advancement.

Unless an ultimate AI-driven recommendation algorithm emerges, the intellectual arms race between fraudsters and defenders in click and ranking manipulation will never cease—only escalating through loophole exploits and fixes.

To maintain AdSense’s trustworthiness—convincingly declaring “this is the most effective ad click, trusted by advertisers worldwide”—would require algorithmic maintenance and development resources at least tenfold those Taobao invested in anti-fraud.

Gu Cheng could provide the foundational framework and project indicators for this algorithm but lacked a ready technical team skilled enough for ongoing development. This required a top-tier research group specializing in search engine matching algorithms, not merely a bunch of hired programmers.

Otherwise, companies like Baidu and Google would have already flooded the market.

The other two essential components were a mature search engine and a vast database of search behavior data.

Given these criteria, in 2002, Gu Cheng’s realistic partners worldwide were limited to five: Google, Baidu, Naver, Yahoo, and Yandex.

Gu Cheng had no desire to collaborate with Americans, Koreans, Japanese, or Russians—only Mr. Li remained.

After digesting Gu Cheng’s visionary plan, Mr. Li swallowed hard and asked, “How did you come up with such a wildly imaginative system?”

Gu Cheng deflected, “That’s not the key. Developing this algorithm in the U.S. would cost at least $20 million. For us Chinese, a budget of 200 million yuan is about right. Continuous upgrades before profitability will require similar funding. I’ll invest 400 million yuan and want 10% of Baidu’s stock in return.”

“Then you’ll set up a subsidiary named AdSense. We collaborate on development—Baidu provides staff and engine, I provide funds and strategic guidance.”

Mr. Li hesitated, “Aren’t you worried I’ll master your idea, raise funds elsewhere, and cut you out?”

Gu Cheng was candid, “Of course. That’s why I only want 10% of your parent company’s shares. Otherwise, I’d ask for more. If you refuse, I can sell the idea to Naver. They might pay well, and since Naver’s smaller, I could have more influence there.”

“Also, I doubt your original investors would increase funding on your terms alone. Everyone can see this strategy’s potential.”

Gu Cheng had carefully calculated the price.

Although historically AdSense was remarkable, becoming the global internet advertising intermediary, that achievement was thanks to Google’s hand.

If Baidu took the helm, lacking search data accumulation abroad and engine penetration, it wouldn’t reach that stature.

Google launched AdSense mid-2003; Baidu only caught up around early 2006. Google eventually captured over 80% of the global ad intermediary market, while Baidu held about 10%.

Even if Baidu launched AdSense six months ahead, it would at best secure 30% market share globally, with Google still dominating over 60%.

In other words, Gu Cheng’s idea could sway about 10% of the global internet ad intermediary market—not enough to make Baidu the global leader.

Still, this market share would more than double Baidu’s presence historically in this field, potentially allowing it to dominate Asia.

Mr. Li and Gu Cheng shared this realistic expectation.

Some things—like algorithms and concepts—aren’t protected by law, so it’s venture capital that buys time.

After pondering, Mr. Li realized that even if he parted ways with Gu Cheng, raised funds cheaper, Gu Cheng might still sell the idea to others, competing with him.

Without Gu Cheng’s strategic guidance, Baidu couldn’t compete against him combined with any of the other four major search engine companies.

Thus, cooperation was the best path; the rest was merely negotiation.

“You invest 400 million yuan for 5% of Baidu’s stock. But we can form a joint venture AdSense company, with Baidu holding 80% and you holding 20%. How about that?”

Gu Cheng quickly calculated: this valued Baidu at 4 billion yuan, and the new AdSense company at 1 billion.

“Fine, I’ll give you some face. But I have two additional conditions, no equity involved.”

Mr. Li relaxed upon hearing this and said, “Go ahead.”

“I want to create a product called ‘YY Tieba.’ Baidu must provide a ‘Tieba’ category widget on its primary search page to direct traffic to YY Tieba. Plus, Baidu must promise never to operate its own Tieba product in the future.”

“Just a traffic funnel? Sure, I agree. I have no plans for Tieba anyway,” Mr. Li replied without hesitation.

Seizing the moment, Gu Cheng added, “Also, I want a copy of Baidu’s four-year database on user search frequencies and results for categories like music, film, gaming, and literature. This will help me sketch national interest patterns for market research.”

Mr. Li grew cautious, “What for? Planning your own search engine?”

He wasn’t foolish enough to nurture a competitor.

Gu Cheng smiled innocently, “No, no, just a full-category content e-commerce. I need in-site search and Baidu’s big data to roughly understand user interests, that’s all. Nothing else.”

Mr. Li considered and agreed, “Happy to cooperate.”

Gu Cheng relaxed, “Great. After completing these, next month, prepare a progress report and accompany me to lobby a government official.”

“A government official? Why?” Mr. Li asked, puzzled.

Gu Cheng briefly explained his idea to leverage Cardinal Qi to pressure China Unicom’s leadership.

Mr. Li was even more baffled, “To involve such a high-ranking official for this public-benefit project? Don’t you know favors like this are valuable?”

Gu Cheng shrugged helplessly, “I have no choice. I treat this country’s internet industry as my own family’s. Anything good for Chinese media is good for me.”

His final bold declaration left Mr. Li utterly bewildered.

Was this a declaration of sovereignty? The emperor as the state itself?