400 Million App Users, 1 Million Phones in Three Years: How Hard It Is to Sell Your Fans Anything

When the core business makes no money, the company starts telling a more complicated story about how it will. That familiar startup path may no longer work.

Judged by revenue, Meitu — maker of 23 apps — is now a phone company.

On Monday night Meitu, which is preparing to list in Hong Kong, published its financial figures. Revenue for the first half of this year was 585 million yuan. Phone sales accounted for 95.1% of it; internet services for 4.9%.

Getting the young women who touch up their faces with Meitu XiuXiu, SelfieCity and BeautyCam to pay for a Meitu phone is the company's most important attempt to make money from its huge user base.

Three years and five phone generations later, Meitu has indeed turned phone sales into its main source of revenue. In 2013, according to the filing, the Xiamen company's biggest customers were still advertisers. Today its top five customers are all distributors that take its phones to market.

Phones look big because internet services earn so little

A high share of phone revenue does not mean Meitu is doing well. Against that 585 million yuan in revenue, the company lost 1.2 billion yuan over the same period once preferred-share valuation losses are stripped out, and carried 9 billion yuan of total liabilities — 3.65 times its assets.

Leave the phones aside. The 446 million people worldwide who open a Meitu app every month brought the company 28.6 million yuan over the past six months. That works out to one fen per active user per month — one hundredth of a yuan.

The apps are not the problem.

Meitu XiuXiu has been a default install on Chinese smartphones since the day it launched. Photo editors such as Spring and Zuji flared up and disappeared; Meitu XiuXiu is still one of the most complete editors on a phone.

The apps that followed each grabbed a different need. BeautyCam sells "one-tap beauty" for plain selfies; MakeupPlus sells "one-tap makeup"; SelfieCity sells city-themed filters.

Since 2014, at least four Meitu apps have held places in the top ten free photo and video apps in Apple's Chinese App Store.

And in livestreaming, the hot market of the moment, Meipai reportedly draws 141 million users a month who spend an average of 31.5 minutes a day in it.

None of this has produced money. Internet-services revenue in the first six months of this year was 31.5% lower than a year earlier.

Meitu's users are women, and advertisers are supposed to be eager to reach women online.

But in the chain that runs from selfie to shared photo, the place best suited to an ad does not belong to Meitu.

People shoot and retouch with Meitu's products, then post to WeChat and Weibo. The advertising value is created not while the picture is being made, but on the social platforms where it lands.

Advertisers, in other words, would rather show you an ad in the Moments feed where you look at photos than while you are slimming your own face in one.

Meitu never built the billboard. Open Meitu XiuXiu, or any other Meitu app, and you will see almost no brand advertising. Most slots that could carry display ads point instead at downloads of Meitu's own apps — Meitu XiuXiu sending you to the App Store for SelfieCity, or SelfieCity handing you off to the short-video community Meipai.

Whether it tried and failed, or simply decided phones were the easier money, Meitu turned to hardware in 2013. That was the new road it chose.

But selling a phone to an app user is hard

When the first Meitu phone, MeituKiss, came out, CEO Wu Xinhong told interviewers that it could be understood as "a selfie machine that makes calls." He had no intention of serving people who dislike selfies; the phone was built for young women who love them.

"We know what kind of beauty effects Chinese girls like. We even know what girls in Korea, Japan and Southeast Asia like, and we tune the effects to their tastes. Our competitors can't do that." Both Wu Xinhong and Yi Wei, who ran the phone business, came back again and again to this one competence: knowing how to define beauty.

Starting with a first generation built by a contract manufacturer under Meitu's label, the company assembled a phone team of more than a hundred people. Making phones is also why it began raising money faster.

Meitu's funding history

2009.1

Seed round of several million yuan from Xiamen Longling Investment

2011.3

Pre-A round of several million US dollars from IDG and Sina Weibo

2013.10

US$5 million Series A from Sinovation Ventures

2014.1

US$55 million A+ round from Qiming Venture Partners and IDG Capital

2014.5

US$114.5 million Series B led by Tiger Fund

2015.6

US$190 million Series C led by ChinaAMC

2016.4

US$136 million Series D led by Jiaoxing Global, at a valuation of US$3.97 billion

2016.8

Meitu files to list in Hong Kong

A phone is a completely different business from an internet service. Research, components, assembly, distribution — every stage costs real money.

Between 2008 and 2012, the five years before it made phones, Meitu raised two rounds totalling a few million US dollars. In the four years since it started making phones in 2013, it has closed five rounds worth more than US$500 million.

Meitu's phones, generation by generation

2013.5

First-generation MeituKiss, 2,199 yuan

2013.11

Meitu 1S, from 1,599 yuan

2014.1

Meitu 1C, aimed at students, 1,399 yuan

2014.4

Meitu 2, from 2,199 yuan

2014.11

Meitu M2, 2,199 yuan

2015.4

Meitu M4, 2,199 yuan; the Hello Kitty and Doraemon editions at 2,499 yuan

2015.11

Meitu V4, standard version 3,499 yuan, Elegance edition as high as 6,099 yuan; launched alongside it, the M4s and two cameras, the Meitu SNAP1 and the Meitu BF1

2016.6

Meitu M6 and V4s, the M6 from 2,399 yuan and the V4s from 3,499 yuan

From launch through this June, the filing says, Meitu sold 953,000 phones, 289,000 of them in the first half of this year.

The numbers also show that Meitu's real average selling price — what the company actually collects per phone — has stayed around 1,500 to 1,600 yuan, only a little above Huawei, OPPO and vivo, and nowhere near the room Apple has to make a profit.

Meitu wanted the selfie to be the thing that set its phone apart. Yi Wei, vice president of Meitu Mobile, told QDaily in an earlier interview that the marketing had pushed the selfie from day one, so that anyone thinking about beautified selfies would think of a Meitu phone. Rival brands, he said, would need time to build an association that strong.

Zhang Wei, Meitu's CTO, added that because the phone is a custom product, Meitu can load bigger and more powerful editing software onto it — software that app users would not want to download at a hundred megabytes or more.

In the end, almost nobody paid for that deeply customised camera. In three years, the 446 million people who use a Meitu app each month bought 953,000 Meitu phones between them: a conversion rate of 0.2%, and that is before subtracting the people who bought more than one generation.

Meanwhile China's smartphone boom has ended. Sales across the market have stopped growing, and building a new brand will only get harder.

Every phone maker now sells its camera and its beauty filters. vivo hired Song Joong-ki to advertise the soft-light selfie mode on its new flagship, the X7. Even the Meizu Blue Charm E, a thousand-yuan phone, makes a point of its smart skin-smoothing mode.

Meitu named "a large share of revenue from smartphones" as one of the three main risk factors in its prospectus, warning that rising component costs could eat into revenue. A maker shipping fewer than a million phones a year cannot compete with the leading brands for suppliers and factory capacity.

The larger risk is that a phone business like this cannot hold up the multibillion-dollar valuation Meitu has reached after so many funding rounds.

The more complicated the business logic, the less it can be trusted

When the users a free service attracts do not produce enough revenue, an internet company tends to tell a more roundabout story about making money. Meitu's version was to sell phones to the people who use its apps.

Even the most successful photo-editing app could not talk enough of its users into buying its hardware.

The most valuable technology companies in China and the United States all make money in plain, direct ways. Apple sells hardware. Google and Baidu sell search ads. Facebook and Tencent sell social ads and games.

Over the past few years, though, the stories that tech startups with big user bases and no income brought to investors have been elaborate.

Sogou's "three-stage rocket," proposed in 2011, ran input method to browser to search engine: the more people used Sogou's input method, the more would use its browser and its search engine, and the advertising money would follow.

For some companies three stages were not enough. "Ecosystem chemical reaction" now appears on roadside billboards.

A closer parallel is Xiaomi. It sells enormous numbers of phones at almost no margin, and wants to sell internet services to the people who bought them.

Xiaomi founder Lin Bin described the original plan this way: "Start with the open-source Android operating system and get the user experience right. Once users accept the OS, build phones, sell them through e-commerce, and in the end make money from software and internet services."

By 2015, when Xiaomi tried to make that turn, it was not going smoothly. Figures obtained by Reuters tell much of the story: Xiaomi's 2015 revenue was US$12.5 billion, but services brought in only US$564 million, more than 40% short of the US$1 billion internal target. Phone sales missed their internal target too — by less than 20%.

Small companies do the same thing. The maker of the weather app Moji Weather, unable to earn from forecasts, tried building a community where people shared weather, and tried selling air-quality monitors.

Now Moji has gone back to the most direct method and put more ad slots in the app. Xiaomi has announced a "new Chinese goods" strategy and wants to sell more hardware built around how young people live.

Meitu is adjusting too. In its listing documents it says it will launch a social e-commerce platform in the first half of 2017, selling fashion, cosmetics and skincare.

Some of that has already started — from Meimei Xiaodian, a nurturing game released in April, to BeautyCam. Meitu is working with makeup brands, moving from virtual try-ons to selling the products outright.

Advertising is back on the agenda as well, written directly into the documents for the Hong Kong listing.

After three years of making phones, Meitu has picked up the most direct way of making money again.

Cover image from Meitu phones