Two Years and 1 Million Apps Later, WeChat Mini Programs Still Aren't a Platform

They are more features for WeChat.

In the space of six months, Miao Zhuang's business selling goods started over twice, both times chasing mini programs.

Miao Zhuang had run after a few startup fads in Zhongguancun. Early this year he rode the concepts of "dimensional-reduction strike" and "new retail" to an investment, then went home to Zhengzhou and opened a six-store fresh-fruit chain called Chao Bianli, or Super Convenience.

In May, with Chao Bianli already losing money, he launched a WeChat mini program called Pin Guoguo, hoping to move his walk-in customers onto it so they could order whenever they wanted.

"We just stuck a QR code up in the store — you scan it and I give you a little gift," Miao Zhuang told QDaily. It didn't turn things around. Chao Bianli closed two months later, still in the red. His own conclusion: people willing to walk in and pick through the fruit themselves mostly don't need it delivered.

In September, back in Beijing, Miao Zhuang built a second mini program, Zan Wu Ji. This time he skipped the physical store and pulled customers in through Moments, selling fruit and household goods. He now takes more than a hundred orders a day at 30 to 50 yuan each. He isn't making much, but he has stopped losing money.

From dressing a business up in fashionable concepts to plain old WeChat commerce: that shift is roughly a miniature of what happened to mini programs themselves.

On January 9, 2017 — the tenth anniversary of the iPhone's launch — Zhang Xiaolong, the father of WeChat, stood in front of the cameras and formally unveiled mini programs. He talked his way from the personal computer to the iPhone and then to mini programs. Mini programs were meant to be a revolution on the scale of the PC and the iPhone.

The revolution lay in challenging the idea that an app has to be installed — and in going offline. Mini programs, Zhang said in his talk, would center on serving offline situations, and the seven scenarios demonstrated on stage included ordering in a restaurant, buying a train ticket, checking bus schedules and finding a nearby convenience store to shop at.

Two years on, a report from the mobile data firm Quest Mobile finds that the mini programs with the highest user activity are in finance, mini games, travel and e-commerce.

That is nothing like the original plan. The whole offline code-scanning scheme Zhang Xiaolong imagined never became reality beyond payments.

Offline, Then Games, Then E-commerce — and Finally the Big E-commerce Firms Tencent Invested In

Mini programs were born just as China's internet industry sank into a collective anxiety about the "disappearance of the traffic dividend." 2017 was the year growth in Chinese smartphone use began hitting its ceiling: the user base stopped growing, but every company still wanted the attention on your phone. The cost of acquiring a user online started to soar.

When website growth flagged, internet companies had mobile apps come along to save them. Now that apps are hard to grow too, WeChat — which nearly everyone in China carries — looks set to hand over its own users to prop up a new platform.

At first mini programs had almost no entry points online. WeChat had no mini program store, allowed no push notifications, no sharing to Moments, no games, and no mini program with the same name as its own official account. Even searching for one required typing the full name exactly before it would appear. Using a mini program essentially meant scanning a code offline.

That was not what developers had hoped for, but nobody dared drag their feet: WeChat had not seemed to get the big calls wrong. It became a national app without buying ads or working the streets, and in three Spring Festivals of red-envelope warfare it caught up with more than a decade of Alipay's work on payments.

But as Miao Zhuang found, opening a shop, getting customers to scan a code and expecting them to keep spending inside WeChat afterward was too optimistic. Enthusiasm for mini programs faded fast.

After the offline approach failed, WeChat opened up a series of online entry points:

Mini Program Changes: A Timeline

2017.03

Official accounts can put mini programs in their menu bar

2017.04

Mini programs get their own circular QR code

2017.05

WeChat Lab launches Search and Top Stories

2017.08

Group chat info pages get a mini program entry point

2017.09

Search can turn up mini programs offering local services

2017.12

Mini games are added, allowing game development; mini program pinning is removed, and a mini program taskbar is added

2018.07

Mini programs can carry ads

2018.09

"Direct to Function" launches: search a keyword in WeChat and mini programs offering the related service start showing up first

What WeChat itself cared about most was still games. Jump Jump, released in December 2017, became the hottest internet product of the year's end because friends competed on it.

Then game license approvals stopped abruptly in March this year, and the whole industry took the hit. Mini program games still don't need a license, but under regulation like this, expecting them to have a future is too optimistic as well.

E-commerce became the biggest hot spot for mini programs. LOOK, a mini program e-commerce operation, signed fashion bloggers including gogoboi and Shiliupo Baogao and embedded mini program product pages in official account posts to sell goods. By March 2018 LOOK had raised more than $27 million across two rounds.

In January this year Mogujie released figures showing that its mini program had accumulated more than 90 million users, with "a conversion rate twice that of the app, 35% next-day retention among active users, and orders placed 2.28 times as fast as on the app."

The peak was Pinduoduo's listing in July. By getting WeChat users to rope in friends for group-buy discounts on cheap goods, Pinduoduo went public three years after it started, its market value on listing day already two-thirds of JD.com's — and now roughly 80% of it.

Miao Zhuang was a devoted follower of the mini program e-commerce story, a believer that social commerce means "fission": one customer telling ten, ten telling a hundred. What he remembers best is a batch of rock-sugar mandarins he brought in, five jin (2.5 kilos) for 30 yuan. Without spending a cent on marketing, he sold more than 500 orders in a day, most of them arriving through mini program links friends had shared.

Spending less to get customers is exactly why mini programs drew so much attention to begin with.

New users are harder to find, and selling online has slowed too. China's total e-commerce transaction volume for the first nine months of this year rose 27% year on year — the first time growth has fallen below 30%.

With selling more goods getting harder, Alibaba would rather talk to stock market investors about the growth of its own revenue. Alibaba's biggest source of revenue is the advertising fees it charges Taobao and Tmall merchants, and that revenue grew 48% in 2017, while transactions across its e-commerce business grew only 27.95% over the same period.

In 2015 Alibaba took in 37.5 billion yuan in advertising against platform transaction volume (GMV) of 3.092 trillion yuan, which means merchants spent an average of 1,213 yuan on ads for every 100,000 yuan of goods sold. A year later, selling the same 100,000 yuan of goods cost 1,391 yuan in advertising. By 2017 the figure had jumped to 1,657 yuan.

Taobao's advertising fees pushed some small merchants toward WeChat. WeChat, in truth, was not ready to take them.

Two years after the launch, every step of shopping inside WeChat is still unfinished.

E-commerce begins with letting people know where the goods are — search, for instance. Starting this summer, WeChat search could look up products directly, but not every keyword finds a matching one.

In September, searching "mooncake" in WeChat turned up Ganso's WeChat store and Yibaizhen on a page filed under "Mooncake — Product." Only a handful of keywords — mooncakes, flowers, skirts, trousers — produced a "Product" result at all, and most of what appeared could be looked at but not bought. Search something slightly off the beaten path, like "Bluetooth earphones," and all that came back were items from JD.com.

From November the "Product" category disappeared altogether, and the only goods left to search were things sold on JD.com — and not even everything JD sells shows up in the results. A PS4, hardly an obscure object, cannot be found.

"We tried every single place in WeChat where you can advertise," complained Jeff, who launched the sporting-goods mini program store Jishangou this July. His grievances ran on many levels: an official account ad is confined to one line of headline and one thumbnail; users in a social setting resist ads, and in group chats and Moments they often look without clicking; a mini program cannot jump straight to an official account and needs a second confirmation from the user, which loses customers; WeChat's ads carry nothing like Taobao's data on what users buy, so precise targeting is impossible…

After a few months of trial and error, Jishangou's click-through rate on WeChat ads was a mere 0.2%, and it gave up. Its ad money now goes to Douyin and Toutiao, and it pulls people from there to order in its Taobao store.

Advertising was only the first problem.

Finding a product and then, instead of buying it outright, putting it in a cart brings fresh trouble. WeChat launched a shopping list in September, buried very deep: a buyer sees the "My Shopping List" entry only after adding to a cart or placing an order at a Youzan mini program store that has the "WeChat Shopping List" capability turned on, so that product and order data have been synced.

Without a single unified WeChat cart, whatever a user drops into an individual mini program's cart is forgotten unless they remember to go back and order. When a shopper hits something they need to think over, it is hard to toss it in a cart and buy it later, the way they would on JD.com or Taobao.

After the purchase come the problems of shipping and customer service. A merchant opening a store on Taobao can post a deposit, collect ratings and use the store's age to show it can be trusted. Even Pinduoduo, deep in its counterfeit-goods scandal, requires merchants to put down a deposit and runs a fast-refund mechanism. WeChat has none of this, and no official customer service either. In Jeff's words: "You buy something, and you don't even know whether I'll ship it."

On November 19 this year ofo, caught in its own deposit scandal, published a post on its WeChat official account selling honey of dubious origin. ofo deleted the article quickly, but a merchant who wants to sell goods only has to change the name and find somewhere else to sell.

The head of Youzan's customer success team told QDaily in an interview that "what WeChat social commerce lacks most right now is a purchase guarantee. Between friends, the conversion rate can reach 20% to 30%; between strangers it is under 1%. That is what the absence of a guarantee does."

WeChat itself has never managed to launch a guarantee system for mini program e-commerce. Youzan launched its own guarantee service this year, which it says lifted the stranger conversion rate from 1% to 3-5%.

But the ones that actually sell things on WeChat are still the big companies and the big brands.

Figures from iResearch and Sinolink Securities' research institute put e-commerce transactions on the WeChat platform in 2017 at close to 500 billion yuan. Youzan, the largest mobile retail service provider inside WeChat, handled 17 billion yuan across all networks in 2017, just 3.4% of that. Pinduoduo handled 140 billion yuan over the same period, and doubled it the following year.

A small merchant's chance of being seen inside WeChat is nowhere near that of a company Tencent has invested in. Search "earrings" in WeChat, for example: the top result is a product page from JD.com, with description, photo and price; second is a Moments post mentioning "earrings"; third is an official account article containing the word; and only fourth is a mini program with "earrings" in its name, showing a single store, by name, rather than any particular product.

The reason is not hard to work out. Tencent holds 18% of JD.com, of Pinduoduo and of Mogujie.

Even among the big e-commerce companies Tencent has invested in, only the biggest get a share of the traffic.

On the eve of its listing, Mogujie halved its fundraising target, from $500 million to $200 million. On its first day of trading the stock fell more than 17% at one point, and over the following days it dropped below the $14 offering price. Its market value is not even a rounding error next to Pinduoduo's.

Both are e-commerce companies 18% owned by Tencent, and both hold a slot in WeChat Wallet. But Mogujie's revenue fell while its own marketing spending rose 7.9% from a year earlier — building a mini program did not get it cheaper traffic.

If Pinduoduo's listing in July marked the start of the mini program e-commerce craze, Mogujie's listing in December marked the moment doubt broke out.

Investors who were still chasing mini programs in the first half of the year have turned pessimistic too. Several from ZhenFund and Panda Capital told 36Kr in recent interviews, flatly, that the valuations investors handed out at the time were a bubble: mini programs may not need installing, but the share of users who open one again the next day is often half or a third of an app's.

New Rank counts twice as many mini program financing deals in the first half of this year as in all of last year, with at least 1.4 billion yuan staked on mini program entrepreneurs. By the second half, record-breaking mini program e-commerce funding news had largely stopped appearing.

Two Years On, 1 Million Mini Programs Have Mostly Built More Features for WeChat

In November 2018 Pony Ma handed out mini programs' report card: 200 million people open a WeChat mini program every day, there are more than 1 million mini programs, and they cover more than 200 industry subsectors.

One million is a big number, though mini programs are easy to build — basically any company that can build a website can manage one. WeChat official accounts are easier still to start, and only 3.5 million of them are currently being updated.

What the 200 million users show is less clear. Meituan has more than 300 million customers, and any merchant link they share with a WeChat friend automatically becomes a mini program link. Someone who never deliberately opens a single mini program becomes one of the 200 million the moment they open one Meituan share link.

That doesn't make these automatic mini programs meaningless. They are far faster, and far nicer to browse, than opening a web page inside WeChat used to be. For WeChat itself, that is an upgrade worth having.

Over the past few years WeChat has essentially become an internet of its own. Inside that closed ecosystem you can socialize, shop, read news, read books, listen to music and play games.

Mini programs make everything you do inside WeChat a little more convenient, and WeChat decides which of them appear in front of more people.

In Aladdin's Top 100 WeChat Mini Programs, November 2018, the top ten were Tongcheng-Elong, Pinduoduo, Meituan Shopping, JD Shopping, Bushubao, Jump Jump, Hot Weibo, Xiaoniangao+ and The Pirates Are Coming — a lineup tilted even further toward big companies than the App Store charts.

Toward big companies Tencent has invested in, especially. Most of the Top 100 were games (31), tools (21), online shopping (19) and content and news (7). Twenty-eight of them had taken Tencent money. Every mainstream e-commerce platform Tencent has invested in is on the list; only 7 of the e-commerce mini programs sit outside the Tencent camp.

Mini programs don't need installing, but they don't need deleting either, and with no push notification of any kind, you use one and drop it. They become individual features inside WeChat, and they can never compete with WeChat itself for time.

Even the companies Tencent has invested in are balancing mini programs against their own apps. Didi still does not offer the basic function of messaging your driver inside its mini program. Since going public, Pinduoduo has been pushing WeChat traffic toward its own app, luring mini program users over to play games and win free fruit.

Open a product page in JD.com's mini program and it prompts you to jump to the app — a privilege of its own. By Richard Liu's own account, the share of JD's new users coming from WeChat fell from 50% three years ago to a quarter last year. JD's earnings reports would also rather emphasize that 85% of this year's orders came through JD's own app, 5 percentage points more than last year.

Every company that wants to make itself a super app has started copying the idea. Baidu's and Alipay's mini programs have both formally launched, and Toutiao and Douyin began closed testing of their own in September and October.

But whoever builds them, and however they are built, mini programs end the same way: they add more features to the super apps for a limited return, and keep people inside WeChat, Toutiao or Alipay.

More than a decade ago Facebook, fresh from a large investment by Microsoft, launched a platform program of its own that let other companies build web apps on the social network — the various crop-stealing games that later swept China were straight copies of one such Facebook app. But after Bill Gates looked at the financial numbers, he told the head of Facebook's platform business: "That's bullshit. This isn't a platform. A platform is when the economic value of everybody that uses it exceeds the value of the company that creates it. Then it's a platform."

Cover image / Visualhunt