Alibaba Has Changed What It Counts Again — How Do the Biggest Internet Companies Tell Stories With Numbers?

In the end, what story a company can tell well still depends on how the product itself performs.

On the second day of June, a market capitalization of $313.6 billion put Alibaba among the world's ten most valuable companies.

The following week happened to be Alibaba's annual Investor Day. Along with a set of high-growth forecasts, the company announced a change in the numbers it would disclose from now on.

Counting how many people buy something on Tmall and Taobao each month, Alibaba said, was no longer enough to show how healthy it is. In its place came "consumers" — not only people spending on Tmall and Taobao, but people using the UC Browser and Youku Tudou.

In the previous fiscal year, 454 million people bought something on Alibaba's e-commerce platforms. Alibaba has not yet published its annual consumer count, but Youku Tudou and UC both have large user bases, so it is easy to guess the new number will come in well above 454 million.

From here on, it will appear in every quarterly earnings report.

Every quarter, a listed company has to publish a financial statement so investors can judge how the business is doing.

The numbers in that statement, the wording released alongside it, and the views executives put out on various occasions together make up a company's story about its future.

The stock market is a place that looks at a company's future. Whether these numbers and the story around them convince investors decides whether the share price rises or falls, and so decides how much capital the company can mobilize to attack its competitors.

Beyond the basics that exchange regulators require — sales, profit and loss — a listed company can also choose to disclose other numbers and build a story around them.

Before internet companies, that was hard. A carmaker is unlikely to talk investors out of watching sales volume and into watching how often people drive the cars they bought.

Internet companies have more numbers to offer, and more ways to explain how their business is going.

Alibaba is not alone. Almost every one of the most successful internet companies picks its numbers carefully, to tell a story that suits it.

Alibaba Has Changed Its Yardstick for the Second Time

"GMV should never be the core standard for e-commerce; commercial infrastructure is what's core. … It's like PV (Page View), which was once a very popular metric among internet companies but isn't what we should really value." That was Jack Ma, speaking to investors at Alibaba's investor meeting last June.

GMV means gross merchandise volume, the value of transactions. If Apple's flagship store sells a 5,888-yuan iPhone, that produces 5,888 yuan of GMV, no matter how much of it Alibaba actually takes.

Two months later, when Alibaba published its financial report for the quarter ending in June, it stopped disclosing GMV. Announcing the results on its official Weibo account, it added: "Do you still think Alibaba is just an e-commerce company? Plenty of investors don't see it that way anymore."

With that, the storytelling standard Alibaba had used for ten years changed.

Ever since Taobao, it had measured itself by GMV.

Alibaba did not invent the standard. eBay, the auction site Taobao first modeled itself on, had long used transaction value as its story. At first the term was GMS, gross merchandise sales; later it became GMV. By some accounts, the reason was that "sales" is the same word many retailers use for revenue, which risked misleading people, and eBay made the change under pressure from the U.S. Securities and Exchange Commission.

Taobao launched in 2003. The platform, which let anyone buy and sell online, aimed straight at eBay, which had entered China not long before. The difference was that Taobao let sellers list goods for free, promising no fees for three years. When the three years were up, Alibaba extended the free period by two more.

Naturally, Alibaba measured its own health by GMV too. For a long stretch, after all, Taobao had essentially no revenue. What revenue came later was the fees shops paid for more exposure on the page — Taobao Zhitongche, Taobao Ke and the like. In plain terms, advertising.

The story ran for more than a decade. It also joined up with Jack Ma's much-repeated line about making it easy to do business anywhere. Taobao then became a platform, running all sorts of online and offline shopping festivals to pull merchants in. But right through to Tmall, which admits only corporate sellers, Alibaba's revenue still came from platform advertising fees, service fees and commissions — not, as with most brick-and-mortar retailers, or Amazon and JD.com, from buying goods cheap and selling them at a markup.

In the prospectus for its 2014 U.S. listing, Alibaba wrote that its three platforms — Taobao, Tmall and Juhuasuan — together generated 1.5 trillion yuan of GMV in 2013. Every year, in the offices gearing up for Singles' Day, the walls are papered with GMV figures from years past: 57.1 billion yuan in 2014, 91.2 billion yuan in 2015 …

But Alibaba's story of GMV abundance was running out of road.

After the second-quarter 2015 results came out, Wall Street's disappointment peaked: revenue grew 28% year on year and core transaction volume 34%, both the slowest in three years. The share price soon fell below $60, nearly half its best-ever level.

The stock then bottomed out and rebounded, until Singles' Day. After Tmall and Taobao together posted 91.2 billion yuan in transactions, up almost 60% year on year, Alibaba's stock fell again, down as much as nearly 4% at one point and closing off 1.94%. On the 13th it fell another 1.37%.

Wall Street's distrust came from doubts about Alibaba's transaction data, and from worry that slowing growth in China would cut consumer demand.

That May, Alibaba said the U.S. Securities and Exchange Commission was investigating the accounting practices on its e-commerce sites around Singles' Day. Also under investigation were the complicated accounting transactions between Alibaba and Cainiao Logistics.

The hedge fund manager Jim Chanos wrote a piece accusing Alibaba of using inflated transaction and user data to make its performance look better than it was.

John Hempton, a fund manager at the Australian hedge fund Bronte Capital, told Business Insider he doubted Alibaba's data: "These numbers are less reliable than China's other statistics," he said.

It was much like what Marshall Meyer, a management professor at the University of Pennsylvania, said about GMV: "Have you ever heard of a logistics company disclosing the total value of the goods it delivers?"

Several Wall Street analysts interviewed by CNBC said Alibaba's model depends on consumption, and that in the long run China's macroeconomic weakness was worth worrying about — enough to cancel out any short-term lift from promotions.

"I believe the next five to fifteen months will be a difficult period for the Chinese economy. There are many reasons, and of course one of them is the anti-corruption campaign, which will definitely have a negative impact." Jack Ma said so himself. As of this year he still hasn't changed his tune.

Amid the chorus of doubt, Alibaba's stock slid back toward $60.

From 2016, Alibaba started making content. Through channels like Weitao (something like Weibo and WeChat official accounts) and Wen Dajia (a consumer community), it invited Taobao shopkeepers and partners to write articles or shoot videos recommending goods, and took a cut of the sales. Through content, Alibaba tried to keep people around, then turn them into the numbers it wanted to talk about.

Weitao works like an official-accounts entrance inside the mobile Taobao app. In it you could find independent media posting the hot topics of 2016 — Wang Baoqiang's divorce, YSL's Star lipstick, Ruby Lin's wedding.

Alibaba did not mind whether the marketing came as livestreams, text or anything else, so long as consumers stayed longer and spent more. For Singles' Day 2016, it brought in hundreds of livestreaming influencers, seen as a more promising way to hold consumers' attention.

Midnight Diner (《深夜食堂》), a talking point these past few days, is something Alibaba tried back in June 2016 — a show made by the Taobao team on "Taobao's Second Floor," themed around Taobao food, running every night from 10 p.m. to 7 the next morning.

In one year, Taobao and Tmall's combined e-commerce revenue rose 45%, to 138.8 billion yuan. It is hard to say whether the increase should be credited to content. Either way, this is something Alibaba has to do.

Because its competitor has shifted from e-commerce to WeChat. Alibaba's new metric, "annual active consumers," is in fact aimed at WeChat — it counts spending on physical goods, services, virtual goods and content, all completed through Taobao, which turns Taobao from a tool for buying and selling goods into something more like WeChat, a more social app competing with it head-on.

At bottom, what Alibaba and Tencent are fighting over is no longer e-commerce, but who gets to monopolize the daily life of Chinese users.

JD.com, compared with Tmall for years, likes to tell a story with little to do with Alibaba. JD.com talks about revenue; Alibaba talks about transaction volume. JD.com follows a different company: Amazon.

Amazon Talked Investors Out of Looking at Its Profit — But It Nearly Died on the Way

On May 31, 2017, Amazon's share price broke $1,000 for the first time, the fifth U.S. stock to do so. Its market value also hit a record $478 billion, fourth in the world and twice that of Walmart, the world's largest retailer.

The difference is that Amazon's 2016 profit was one-sixth of Walmart's. Set against the other four most valuable companies, Amazon falls even further behind.

"Your margin is my opportunity." Amazon's founder and chief executive, Jeff Bezos, coined that line early on.

But profit still matters, whether for Walmart, which sells cheap goods at scale, or Tiffany, with few stores and high prices. For a century companies have treated profitability as the key measure of business success, taking it for granted the way we take gravity for granted: you must produce profit in order to create long-term shareholder value.

The world's leading companies are the ones running profit-focused business models. The very existence of the phrase price-to-earnings ratio is proof of how much profit matters. Investors use it to judge whether a share price is reasonable, usually dividing the share price by annual earnings per share, though you can just as well divide market value by annual profit.

Take Amazon's $478 billion market value and $2.3 billion annual profit: a P/E ratio of 207 (478/2.3). There is a simpler way to read that. Assume Amazon nets $2.3 billion every year from now on; for every $207 you pay, you get $1 of the company's profit, and it takes 207 years to get your money back.

A U.S. stock with a P/E above 30 is usually taken as a sign the share price is padded. Amazon's has reached more than 4,000. A company that loses money has a P/E of zero — and Amazon spent many quarters right on the line between profit and loss.

The story Bezos told was a different one. Amazon signs contracts with suppliers under which it takes the goods first and settles the bill three months later. It then sells those goods and collects the cash. When the three months are up, Amazon repays the principal and keeps what it earned on top. In effect, as long as Amazon keeps selling, it can keep using its suppliers' payment terms to fund its own growth.

There is another point. Between taking cash in and paying it out, even if Amazon did nothing but leave the money on deposit, its several billion dollars of annual revenue in the early years would throw off a lot of interest. In 1999, Amazon posted an operating loss of $600 million and interest income of $45 million.

Plenty of companies have tried to get the world to accept that a business needs growth, not profit — the fallen internet companies Pets.com, eToys.com and the rest. And not only internet companies: plenty of old industries, steel and textiles among them, keep running on bank debt. When the banks stop rolling the loans over and the company's cash chain snaps, bankruptcy restructuring follows quickly. This April, Shandong saw a run of bankruptcies among private companies caused by debt defaults and banks pulling their loans.

LeEco, much in the news lately, is the same story. Jia Yueting failed to keep his cash chain under control, pledged nearly all his shares, and burned 10 billion yuan on the car business, some of it money from other subsidiaries. In the end he burned away his own job as general manager of his listed company.

Amazon was no exception. Getting the world to accept its standard took a fight.

In the 1990s it expanded blindly, investing in Drugstore.com, Pets.com, Kozmo.com, Living.com and others. Then the bubble burst; Pets.com and Living.com collapsed, costing Amazon $300 million.

Ravi Suria, a convertible-bond analyst at the investment bank Lehman Brothers, published a 27-page report in 2000 arguing that, viewed from the bond side, Amazon's credit was very poor and getting worse. "We are convinced that the company will run out of cash in the next four quarters, unless it can conjure up another round of funding like magic," Suria said.

But Amazon was not LeEco. Henry Blodget, editor-in-chief of Business Insider, saw as early as 1998 that Amazon was different. He predicted the stock would break $400 within the year, praised what Amazon had achieved in fulfillment, and argued it had already extended its franchise beyond traditional books and music, with the number of internet shoppers set to rise sharply before long. When he said this, Blodget was still an unknown securities analyst.

When the bubble arrived, Amazon pulled in: it stopped the blind acquisitions and the infrastructure build-out. A $672 million round of funding it had taken earlier arrived at exactly the right moment to help.

More important was the core business. E-commerce took off just as Blodget had said. Amazon was no longer only a bookstore; it became the place where you could net every kind of product at once. The distribution centers Amazon built while it was flush with cash later cut its distribution and fulfillment costs, which let Amazon clear inventory as fast as any American retailer, which let it keep raising sales even as the economy slowed, and so avoid a break in its cash flow.

In the end Amazon stopped losing money, but it never banked the profit either — the cash went into expansion. Bezos's early years at a hedge fund may have given him a better read on Wall Street's mood than most executives have. "Amazon got the cheapest capital in human history by telling a story," said Scott Galloway, a professor at NYU's Stern school. Profit, he said, is like heroin: once investors get used to a company making money, they find it hard to accept a story about grabbing market share without any.

Doubts came back later — the phone that failed in 2013, for instance. But then cloud computing, after years of investment, reached the point of making money, and in the first quarter of 2015 Bezos disclosed its results for the first time. From then on Wall Street knew Amazon had built a second strong business outside e-commerce, one that genuinely brings in billions of dollars of profit a year with almost nobody catching up.

Today Amazon's P/E ratio is not worth thinking about. How much profit it shows has become something Bezos can dial: when the stock is low, cut spending and turn a profit; when the stock is high, raise spending and put the money into future businesses. Round and round.

That makes Amazon hard for its competitors — Walmart, say — to fight. Every time Walmart puts money into e-commerce its profit shrinks, and its share price slides — even though Walmart's profit last year was still billions of dollars more than Amazon's.

JD.com gets called the Chinese Amazon. Its founder, Liu Qiangdong, often borrows Bezos's lines. In a 2012 speech, Liu said: "The more cash flow you have on your books, the more cash you can invest, and the more profit you make."

That is what he said. But on the same day JD.com released its results for the final quarter of 2016, it spun off JD Finance's assets. Liu Qiangdong said this was to make better use of money from China's capital markets; really it was to cut the losses in the listed part of JD.com.

Today JD.com is still trying to convince investors it can stanch its losses, just as Amazon was in the early 2000s.

Google Changed How the Ad Industry Counts

In 1998, in the middle of his buying spree, Bezos met two young Stanford computer scientists, Larry Page and Sergey Brin, and gave them a $250,000 investment. That money, together with later rounds from Kleiner Perkins and Sequoia Capital, helped bring about Google's rise — and changed the whole advertising business.

Digital advertising was born almost the same day as the internet business. In 1995, when The New York Times interviewed Jerry Yang, who had founded Yahoo a year earlier, he was already thinking about making money by showing ads on web pages. That became Yahoo's business model. The banner ad we all know was born around then. Open a page and it might sit anywhere — top, middle or bottom.

For advertisers, the format worked poorly. A banner simply appeared on the page; you had no way of knowing how many people saw it, and so no way of knowing what tens or hundreds of millions of dollars of ad spending actually returned.

Advertising is not counted that way now. Most ads charge by the click. Under that model, however many times an ad is shown, the advertiser pays only when a user clicks. Each click hands the advertiser a piece of information about that user. Collect enough of it and you can sketch a profile of the potential audience.

All of this started in 2002.

Google had just turned four. It had begun selling ads two years earlier, but in the same way Yahoo did. In 2002 a new approach arrived.

Today, in every set of results, Google discloses how the amount it earns per ad click is changing.

It can do this because, as a search engine, it knows exactly what users have been looking for. A man searches for razors; a woman searches for shoes, bags, makeup — and the matching ads Google shows them are likelier to draw a click.

In April 2005, Google announced it was buying Urchin and would make a service that had cost money free. This was the ancestor of the website traffic service Google Analytics.

Google opened it to the market free of charge, deliberately making the market more and more transparent — and collecting more user information for itself in the process.

Analytics crashed the moment it went live. In the first 48 hours nearly every Google server went down, unable to handle the wave of data pouring in. Eric Schmidt later called that crash Google's most successful disaster. For nearly a year Google had to ration access, before finally opening the service to everyone.

After releasing the analytics tool, Google's whole business model turned on one thing: getting users to click more ads.

It built its own browser and gathered cookies everywhere to know what users had looked at. When a user visits another page, the browser sends the cookie to the server, so the server knows what they had been shopping for; if they go on to shop for a drink, the server appends the new product information to that same cookie.

Google succeeded in filling third-party websites with ads of every kind.

Once the world accepted Google's standard, its rival Yahoo was quickly overtaken on advertising revenue. China's Baidu copied Google's method more or less wholesale and beat the three big portals too. What Baidu did not copy was how to make it simple and clear to users that something is an ad.

YouTube, Gmail, Maps, Calendar, Photos and the rest sweep up as much user information as they can, refining the ad recommendations further so they fit users' tastes better, and in the end give users a reason to click.

The payoff: Google had steady advertising revenue, did not need to charge for its services, and used good services to keep users. When Microsoft's Hotmail, under cost pressure, started deleting accounts that had not been logged into for 30 days, Google offered Gmail with 5 GB of free storage.

And by using Gmail to link up Google account information, Google let users log into every service with one account, connected to 80% of the world's smartphones, telling it more and more about their personal details and interests.

Only when social networks began swallowing people's time were Google's advertising prospects shaken.

Facebook Equals Social Networking

The second-largest advertising company today is Facebook. Its latest published number is 1.9 billion users a month.

The people who use Facebook each month equal 66% of the internet population outside China. The 1.9 billion figure is also 500 million more than the population of China, the most populous country on earth. And it does not include the services Facebook has bought, such as WhatsApp and Instagram.

Inside this enormous network, its chief executive, Mark Zuckerberg, is treated rather like a president, spending more than $4 million a year on personal security. What makes him look still more like one is that lately Zuckerberg has spent each month visiting different American cities to meet and talk with people from different walks of life, and take photographs like this:

He used to be a completely different person.

"I'm CEO, bitch." Zuckerberg blurted that out at a design team meeting in 2006. The designer on the receiving end later had it printed on Zuckerberg's business cards.

That year, when Forbes interviewed Zuckerberg, the $1 billion Facebook was a two-year-old social network with only a couple of dozen employees.

Some compared it to MySpace. But all Facebook could tell the world then was how many registered users it had: 9.3 million. That was the number Facebook was willing to talk about, to prove how popular it was. MySpace, meanwhile, was already talking about more than 46 million people a month visiting its network.

"Registered users" has plenty of problems; like GMV, it inflates easily. Renren claimed in 2012 that it had passed 162 million registered users. How long has it been since you last used it? Anyone a little younger doesn't even know the site existed.

In 2006, something else that brash CEO pushed through finally let Facebook stop talking about registered users — the site redesign that introduced the News Feed.

Ruchi Sanghvi, an Indian student, was 25 when she joined Zuckerberg's team, and she was in charge of building the News Feed. Until then, when users logged into Facebook they saw their own profile page. The News Feed instead showed a list of status updates from a user's friends — profile changes, upcoming events and birthdays, and other updates, refreshing one item after another.

When the News Feed launched in 2006, users pushed back hard. "When we launched News Feed, we had fewer than 10 million users in total. When Mark announced at a press event that we were launching News Feed, nearly a million users objected to it," Sanghvi said. Zuckerberg held out against the pressure. Later, in 2013, Facebook changed the News Feed algorithm so posts were no longer ordered by time, the aim being to show users more of what actually interested them. That arrangement is standard for social networks now.

In 2008, Facebook began disclosing on its own site how many users opened it each month. Older social networks such as MySpace had launched earlier and piled up more registered users, but few of them talked about "how many people use it each month," so there was no way to show how active the site was.

Facebook talked about it — and was so proud of the numbers that it announced every additional 25 million. The practice did not last long, because it was growing too fast.

Active users are also an important number for selling ads.

In 2009, 360 million people used Facebook each month, 2.45 times the year before. Companies liked buying ads on a platform that was more active and growing fast, because more people would see them. Marketplace, the ad platform launched that year, gave Facebook a system for placing ads at scale and drew still more companies in.

Between 2007 and 2009, Facebook's revenue grew from $153 million to $777 million, 99% of it advertising.

In 2012, eight years old and having raised $2.34 billion in total — a few percent of what Didi or Uber would later raise — Facebook listed on Nasdaq at a valuation above $100 billion. In the prospectus, Facebook also formally adopted "how many people use it each day" as its new measure of engagement. At the end of 2011, 483 million people used Facebook every day.

Facebook had another big problem to solve: moving to mobile. Its prospectus called "mobile first" the company's guiding principle.

It rebuilt its official app, replacing the HTML5-driven web app on iOS and Android with native apps and lifting the experience directly. By user ratings, its score in Apple's App Store climbed quickly from 1.5 stars to 4; by usage data, engagement more than doubled after the new app shipped. For an app with hundreds of millions of users, gains like that do not come easily.

Having changed what people used to look at Facebook, Facebook went on to enrich what people looked at on Facebook. Its partnership with the streaming music service Spotify let users drop auto-playing videos straight into the News Feed. Facebook turned itself into a platform gateway, launching Open Graph and plugging in services such as Pinterest, Airbnb and Netflix. Through that platform, users could find out what their friends were doing, and talk to them about it, without leaving Facebook.

Phones are not PCs. In the PC era, two generations of search engine, Yahoo and Google, were the door people went through into the internet. On phones, social and messaging tools gradually replaced search engines as the thing you use most, and kept you online around the clock.

Daily active users became the key number for judging whether a social service still has room to grow. Twitter, once mentioned in the same breath as Facebook, has barely grown since 2014 — no more than 140 million people open it each day. Wall Street believes Twitter has little growth left; "Twitter is for sale" has even become the best news the company has had in years, sending the stock up 20% the moment the rumor appeared.

As of the end of March this year, 1.284 billion people opened Facebook every day. Other services near the top include WhatsApp with 840 million, Instagram with 200 million and WeChat with 768 million.

Speaking of Tencent: you don't see much data on QQ, only how many people use it each month. Exactly how many use it each day, Tencent will not tell you.

QQ is rather like Facebook six or seven years ago — more of a PC product, not used often enough, so only a monthly user count makes it look big.

WeChat, having said that more than 700 million people use it every day, also tells you it handles 600 million offline transactions a day, and that half its users spend an average of 90 minutes a day inside the app.

In the end, what story a company can tell well still depends on how the product itself performs.