The Email That Made Tencent Build Its Own Replacement

In November 2010, Zhang Xiaolong spent one night studying Kik. Before dawn he sent Ma Huateng an email. The email contained a judgment, not a business plan or risk report. Mobile internet would produce a new messaging app, and that app could hit QQ from behind. If Tencent did not build it, someone else would. Kik had reached a million users 15 days after launch. Its trick was simple. It read the phone’s contact list and let contacts message each other for free. On a PC that was unremarkable. On a phone it pressed on QQ’s weakest spot.

Ma replied the same day. He said yes. The timing was ugly. The 3Q War was at its worst. On November 20, 2010, the MIIT notice ended the fight. That same day, Tencent’s Guangzhou team started the WeChat project. Fewer than 10 people shipped version 1.0 in under 70 days. It launched on January 21, 2011.

Asset allocation was nowhere in that decision. Tencent heard someone picking the lock downstairs at midnight, threw on clothes, and went down.

Mobile QQ was too much QQ

QQ had a mobile version. Why was it not enough?

Ma Huateng addressed it himself. Half of Mobile QQ’s users were on PC and half on mobile. Only WeChat was designed entirely for mobile. WeChat also had no concept of online or offline, because a phone is always online.

Many QQ designs assumed people leave the computer: online, invisible, away, busy; suns and moons earned by staying logged in; offline messages. Those features made sense in the PC era. Moved onto a phone always carried in a pocket, they became a burden.

The relationship chain differed too. A QQ number was an online identity. Many friends were netizens, gaming teammates, classmates not contacted in years. A phone contact list held your boss, your parents, the courier, people in real life. Kik’s breakthrough was exactly the contact list, a relationship chain QQ did not hold.

Mobile QQ failed because it carried too much QQ. It had hundreds of millions of users’ habits, a pile of existing features, and a PC-era interaction logic. Behind every possible change stood users, teams, and metrics. Touch any one and someone pushed back.

Zhang’s side started from zero and could use “less” as a weapon. Ma once worried that a competitor could add features on top of WeChat and claim innovation. Zhang replied that WeChat was already so minimal it had nothing. To surpass it, others would have to add things, and once they added, they could not surpass it. That was a bit of a humblebrag. Given 2011’s mobile networks and phone performance, “light” was the experience.

WeChat took demand QQ had not eaten

A large part of the cake WeChat took was the carriers’ SMS business.

MIIT data: in 2012 national SMS volume was 897.31 billion messages, up only 2.1% year on year, a marked slowdown from 6.2% growth in 2011. During Spring Festival 2013, national voice call volume fell 20% from the previous year. From January to August 2013, national mobile SMS volume was only 506.16 billion messages, down 17.9%. The carriers grew anxious enough that on March 31, 2013, MIIT Minister Miao Wei publicly said WeChat might charge fees, though not by much. Tencent later made clear it would not charge users for basic WeChat services, and the measures were never implemented.

WeChat opened a market QQ had not eaten. Reading it as “the same users being diverted” misses that increment.

The egg basket metaphor fails here. If WeChat had merely moved QQ users from the left pocket to the right, it could not have grown like that. What it grew was the always-on communication demand of the mobile internet, the contact-list relationship chain, voice messages, People Nearby, and the products later layered on top: QR codes, payments, Official Accounts, Mini Programs.

Three teams raced. The email dozen won

Tencent tried to change. Three teams, QQ, QQ Contacts, and QQ Mail, worked on it at the same time. An internal horse race.

The Mobile QQ side was far stronger than Zhang Xiaolong’s. The Mobile QQ and Qxin teams reported directly to Group Vice President Liu Chengmin. Zhang was only general manager of the Guangzhou Research Institute, a big gap in rank, and the former was a team of over a thousand.

The winner was a dozen or so people from an email background.

Two reasons.

On product, old products find subtraction hard. Mobile QQ carried hundreds of millions of users’ habits and a long list of existing features. Behind each feature were users, teams, and metrics. Cut any and someone screamed. Zhang’s side started from nothing and could treat “less” as an advantage.

On organization, Ma later acknowledged the problem. Discussing WeChat’s birth, he said the three products had belonged to different business units, that the company structure was the problem, and that it was adjusted later.

Management theory calls this the innovator’s dilemma: the more successful a big company is with an old product, the harder it is to incubate its replacement inside that product. Tencent’s solution was to build the new thing on a separate site, so the new team owed nothing to the old product’s baggage.

The horse race sounds wasteful. Tencent did it on purpose. Ma said companies often need some redundancy, tolerance for failure, moderate waste, and encouragement of internal competition and trial and error.

Losing to yourself beats losing to Xiaomi

Ma put his position more bluntly. In a speech he admitted that if WeChat had come out of another company, “we probably could not have stopped it at all.”

QQ being hit in the mobile era was almost inevitable. What could be chosen was who hit it. If MiTalk hit it, users and money flowed to Xiaomi. If WeChat hit it, QQ was hurt but users and money stayed at Tencent.

After WeChat won, dozens of apps became cannon fodder: MiTalk, Feixin, similar products from the three carriers. The victims included Tencent’s own QQ. Tencent knew it would wound its own child and fired anyway.

There was no composure of “diversifying risk” here. This was self-displacement. Rather than wait for someone else to replace me, I replace myself first. Losing to yourself beats losing to Xiaomi.

WeChat did not take off immediately

In its first six months, WeChat had fewer than a million users, a number that almost equaled failure inside Tencent. After voice chat arrived in May 2011, daily new users reached 50,000 to 60,000. In July, People Nearby launched. Without using QQ resources, daily new users jumped to 100,000. Zhang Xiaolong said this feature turned things around completely. By March 29, 2012, WeChat had 100 million users, 433 days after launch.

One fair objection is that WeChat won on QQ’s traffic alone.

Half and half. Ma said the real start came from push-to-talk voice, plus contact-list integration and QQ friend data import. You saw a contact had installed it, so you rushed to install it too. The product had to be good first. Only then could QQ’s relationship chain amplify it.

Kik had a good product and no relationship chain, and it never broke through in China. Mobile QQ had the relationship chain and dragged its baggage, and it did not win either. Only WeChat had both.

WeChat’s victory had no single cause. Product judgment, organizational arrangement, timing, and relationship-chain amplification stacked together. Remove any one and the story changes.

Differentiation grew out later

At first Tencent did not intend to divert users. It wanted to migrate QQ users as a whole.

In a 2012 interview, Ma said WeChat’s initial relationships were imported from QQ. Most WeChat logins still used QQ numbers and passwords. He saw the two as additive, not a complete replacement.

Differentiation emerged only later. Ma mentioned at the time that the Mobile QQ product line would compete differently from WeChat. Young people pushed the split too. Ma observed that many kids still used QQ, and that this had nothing to do with technology and everything to do with human nature: they did not want to be in the same circle as their parents. Once Mom and Dad are liking posts on Moments, nobody wants to post an edgy teen signature there.

The latest data shows both baskets still in place. Tencent’s Q2 2026 financial report puts WeChat and Weixin combined MAU at 1.439 billion, up 2% year on year, and QQ mobile terminal MAU at 520 million, down 2%. In Q2 2024, QQ MAU was still around 570 million. The decline is obvious. A base above 500 million still makes it a first-tier social product in China.

The two baskets are an after-the-fact result, and they do benefit Tencent. QQ holds students and young people, and file transfer and group chat still have a good reputation. At the moment the project was approved, though, nobody was acting on a plan to diversify risk.

Do not mythologize the horse race

WeChat’s success makes it easy to mythologize internal horse racing and self-revolution. Hold these facts together.

Tencent lost battles in the same period. In microblogging it entered eight months after Sina and returned empty-handed. Ma himself admitted Tencent did not always seize the right opportunity. E-commerce was one it missed. Sending a few more teams to race does not guarantee a win.

WeChat’s success also mixed in conditions hard to replicate: Zhang Xiaolong’s product judgment, the timing window at the end of the 3Q War, and the instant-messaging technology and relationship chain QQ had accumulated over years. An ordinary big company copying internal horse racing would most likely get three mediocre products and a pile of internal friction.

One more risk gets ignored. Left hand hitting right hand works only if the decision-maker is willing to let the new product eat the old one. With a management team that cared more about current revenue, the Mobile QQ team would most likely have been protected, and the WeChat team would have been told not to conflict with the core business. The story gets rewritten. Commentators have wondered what Tencent would look like today if Zhang Xiaolong had not acted, or if Ma Huateng had not replied to that email. History was one email away.

The market divided the baskets later

Tencent built WeChat because it feared being overturned in the mobile era, so it acted first. It knew it would hurt QQ and still let the new product win, because losing to yourself beats losing to Xiaomi. QQ and WeChat each holding their own territory today is a basket the market later divided for Tencent.

“Don’t put all your eggs in one basket” sounds like calm asset allocation. It does not hold for Tencent at the end of 2010. Tencent then was not calmly doing portfolio management in an office. It heard lock-picking at the door of the mobile internet. It threw on clothes, went downstairs, built a new boat, and moved users from the old boat onto the new one, even if the new boat would ram and sink the old one.

WeChat is the new boat Tencent built with its own hands when the old basket could tip over entirely.

The market divided today’s two baskets after the boat was already in the water.