19: Chapter 19 A Closer Look at the Data on the Plunge

One week after Weikan went live.

Tencent Group headquarters, 12th-floor conference room.

The long table was filled with people, laptops on the table, and a glaring data comparison chart on the projector screen.

Sun Qiang at the head of the table threw the data report onto the table, the papers sliding half a meter across the smooth surface.

Everyone in the conference room lowered their heads, no one dared to speak first.

"It's been a week."

Sun Qiang looked around, his gaze scanning every face.

"Weikan has 10 million daily active users, sounds pretty good, right? But look at the data below—average usage time is less than ten minutes."

"The seven-day user retention rate is not even 15 percent."

He pulled out another chart and held it up in mid-air.

"And what is supporting this 10 million daily active users? Pop-ups on the Tencent Messaging App, recommendation slots on Tencent Video, and the weekly sign-in collaboration with our stable of games."

"To put it bluntly, we are using everything Tencent Group has accumulated over the years as a crutch for Weikan."

"Every time a pop-up appears, millions of users open Weikan."

"But can we keep popping up forever? Every day a recommendation slot is up, it's like using Tencent Video's traffic for a blood transfusion. How long can we keep that up?"

"More importantly."

He flipped to the next page of the chart.

"Today is the last day of the weekly sign-in."

"According to our activity rules, the sign-in event giving away Q coins and skins ends today."

"Users come to Weikan to sign in for the rewards, not to scroll through videos."

"Once the rewards are gone, will they stay? According to our previous estimates, after today, without the stimulation of Q coins and skins, the daily active users will drop significantly."

"What's the estimate?" someone asked.

In the corner, a data analyst, Zhang Li, spoke up cautiously.

"According to our internal most optimistic model, Weikan's daily active users might drop to a fraction of what they were, or even lower."

"Because the pull of the sign-in event was too strong, it masked the users' reaction to the product itself."

The air in the conference room grew colder.

"There is one more thing."

Zhang Li pushed away the laptop in front of him, turned the screen around, and his voice dropped very low.

"Manufacturer complaints."

"We've received more complaint calls this week than all our other business lines combined."

"The advertising conversion rate is nearly an order of magnitude lower than on Douyin. We've been doing advertising placement for over a decade, and we've never encountered such a low conversion rate."

"Several major manufacturers that invested over a million are already asking if they can get refunds. To be honest, if it weren't for the contracts, two-thirds of the manufacturers would have already left."

Sun Qiang was about to speak when his phone on the table suddenly vibrated several times.

It was a link sent by the Public Relations Department, with a headline in bold black text.

"Douyin's official blog latest data: Daily active users exceeded 1.2 million, average usage time is 50 minutes."

Sun Qiang took a deep breath and clicked the link on the computer projector.

Everyone in the conference room saw the image.

Fifty minutes.

On the same day, at the same time node—Weikan's average usage time is less than ten minutes, while Douyin's is fifty minutes.

Zhang Li's voice was dry.

In the quiet conference room, every word was clearly heard.

"We poured in so many resources, our daily active users are nearly ten times theirs, but their user time is several times ours."

"All the users pulled into Weikan by our pop-ups scroll for a few minutes and leave, then turn around and download Douyin, staying on Douyin for an hour."

Sun Qiang flipped his phone face down on the table, his voice dropped very low.

"Someone tell me."

"They are both Short Videos, both scroll up and down to switch videos, even the interface looks exactly the same. Why is the gap so big? What exactly are users seeing on Douyin that they aren't seeing on our Weikan?"

A middle-aged man with glasses sitting in the corner stood up. He was the technical lead of the Weikan R&D team, his face whiter than the report paper in his hand.

"General Manager Sun."

"There is something we have been wanting to report to you, but because we were rushing the launch of Weikan, the whole process was too hurried, and we never had a chance to explain in detail."

"Speak."

"Douyin's videos are not pushed randomly."

He projected an image onto the screen, which was a detailed analysis report their team had worked overtime this week to study Douyin, with dense data curves covering an entire page.

"It has a complete Recommendation Algorithm behind it."

"The system calculates the user's interest tags and behavioral patterns based on every like, every pause, and the time and speed of every swipe, then precisely pushes videos that each user might be interested in."

He flipped to the next page.

On it was a comparison chart of several user behavior paths.

"We pulled dozens of test accounts to conduct a controlled experiment, letting people with different interest preferences scroll through the same initial content on both Douyin and Weikan, then analyzed the differences in subsequent push paths."

"On Douyin, accounts that like basketball scroll through only basketball-related content, and accounts that like makeup scroll through only makeup tutorials."

"But on Weikan, everyone's push path is almost identical; there is simply no distinction in user profiling."

"The gap between the two in this dimension is no longer a question of who is running faster on the same track."

"This algorithm becomes more accurate the more it is used. The more the user scrolls, the better the system understands the user. The better it understands the user, the more the pushed content matches their taste, and the more the user can't stop scrolling."

A few gasps of astonishment sounded in the conference room.

After all, this was the first time they had heard of this concept. Why hadn't anyone in the entire Tencent Group thought of such a perfect algorithm before?

And most importantly.

This big data algorithm was actually created by a college student who hadn't even graduated yet!

At this moment.

Many people in the room looked at each other, seeing quite a bit of shock in each other's eyes.

They still remembered that when Douyin was just launched, Inspur Technology only had three employees.

"You mean..."

Sun Qiang paused.

"We were left behind by Douyin from the very beginning in terms of user experience?"

"It's not a gap in dimensions, General Manager Sun. It's a generational gap."

"Then why didn't we discover this when we were developing Weikan internally?"

Sun Qiang's voice rose by half a degree.

"Because..."

The technical lead swallowed his saliva.

"Because we only spent one week at the time to perfect and build Weikan."

"One week."

"We put all our energy into replicating the interface, making sure every button position followed Douyin's to ensure that the users' operating habits on Weikan were consistent with Douyin."

"But we never spent time analyzing how Douyin's recommendation mechanism works."

"We thought it was just randomly pushing a few popular videos, and whoever was lucky could make it to the popular list."

"We thought the core competitiveness of Short Video was the quantity of content, so we signed a large number of creators and moved a large amount of videos over."

"But we didn't know that without the support of an algorithm, no matter how much content there is, it's just goods piled up in a warehouse, unable to find buyers."

The conference room fell silent again.

The faint sound of the air conditioning vent sounded particularly harsh at this moment.

Outside the window was the afternoon sun of Shenzhen, and the light spots cast through the floor-to-ceiling windows fell on the dark conference table; no one paid attention to it.

Sun Qiang closed his eyes for a moment.

When he opened them again, his voice had returned to calm.

"Two things."

"First, set up a special team for the Recommendation Algorithm, pulling the best people from the Social Business Group and the Video Business Group, without considering the cost."

"Within three months, we must create our own Recommendation Algorithm to use on Weikan."

"We don't have a second chance to make mistakes."

"Second."

He turned to the other side.

"Ask the investment bank to help facilitate, and contact Inspur Technology directly."

"Don't talk about cooperation, talk about acquisition. I don't believe a regular startup company can hold out in the face of enough money."

"As long as we get that algorithm in Qin Mo's hands, Weikan will still have a window of opportunity to make a comeback."

"What we lack most now is not money, but time."

The people on both sides of the conference table nodded at the same time.

No one mentioned that Douyin was not considered an opponent at all, and no one brought up the celebration banquet they had prepared.

They realized in one week where the real threshold of this track was.

Everyone realized.

The software Douyin was by no means as simple as it looked.