Twice in a Week, Didi Says It's Going Global — What's the Rush, and How Big Is the Opportunity?

If Chinese regulation ends up as strict as this month's draft rules, Didi may have no choice but to go overseas.

"We aspire to become a global company."

Yesterday, at The Wall Street Journal's WSJD Live conference, Didi President Jean Liu described the company's overseas strategy that way.

Liu said Didi is weighing which new markets to enter. If it decides the companies already in a market aren't strong enough, it will go in and compete directly.

That was the second time in recent weeks that she had said, in the United States, that Didi would expand overseas — and her tone on competition was firmer than before.

A week earlier, at an event hosted by Vanity Fair, Liu said Didi would definitely go abroad, but that it would "be a sponsor and supporter of local market partners. If there's no local player, we'll do it ourselves."

Until then, Didi had talked publicly only about the Chinese market.

In April this year, Bloomberg Businessweek interviewed Liu and asked how she saw global expansion. What she stressed instead was how big China is: Didi's penetration there was only 1.1%; if it reached 5.5%, ride orders would grow 60-fold — an enormous market already.

Earlier still, all Didi had talked about was working with partners so its users could hail rides directly while abroad. A Chinese user landing in the United States, for instance, could open the Didi app and call a car from Lyft, the local premier-car service.

At the Code Conference in June, Liu said it again: "Only 1% of trips in China are handled by Didi. The room to grow is enormous."

In the internal letter announcing the acquisition of Uber China in August, Cheng Wei and Liu still used mild language about the global market: "After the deal, Didi will continue to work with global partners to connect local resources and give users a wonderful cross-border travel experience."

Over the past month, Didi's attitude toward overseas markets has changed sharply.

Why Is Didi in a Hurry?

Didi has just beaten Uber in China, and now it faces a new problem: less than a week remains before local ride-hailing rules take effect city by city.

Of the 83 prefecture-level or higher cities that have published draft rules for public comment, more than 60% set strict requirements on vehicle models. The cities with looser rules are mostly in less developed regions. Beijing, Shanghai, Guangzhou and Shenzhen — the country's four largest ride-hailing markets — essentially all require B-class cars.

If the final policy doesn't change much before it takes effect, Didi could well end up a high-end taxi company.

After acquiring Uber China, Didi was valued at $38 billion. That is the valuation of a company with a lock on travel for 800 million city dwellers. A high-end taxi company isn't worth that much.

If Didi accepts that outcome, its valuation will have shrunk by the time the policy eventually loosens. And since that valuation is tied to the value of the shares its employees hold, a drop would weaken its ability to attract talent.

Expanding overseas, and making money where regulation is looser, becomes the obvious choice.

But can it go just because it wants to?

How Much Money Does Didi Have?

The model for this globalization game was settled long ago: burn cash. Backing a local startup in a new market or building a team there from scratch — either way it takes a great deal of money.

Didi's most recent investment was $119 million from Foxconn Group. After that round, Didi's valuation reached $35 billion. By Didi's own account, counting that funding plus a $2.5 billion loan from China Merchants Bank, it has at least $10 billion in cash it can call on.

That is roughly all the money Didi can reach in the near term. If regulation in China tightens, raising more at today's valuation won't be easy.

That is because Didi as a whole still isn't profitable. Liu revealed at the WSJD conference that revenue for the year through September had reached $20 billion, but that figure includes taxi-hailing orders that make no money at all; the cut taken on premier-car and express-car rides is only around 20%-25%, and a good share of that revenue goes straight back out as subsidies.

Uber, founded seven years ago, also has more than $10 billion it can draw on. Their valuations differ by half, but on cash in hand the two are evenly matched.

Didi Already Has Some Groundwork Overseas, But the Picture Is Complicated

Last December, the company — then still called Didi Kuaidi — said it would link its product with those of Lyft, Ola and GrabTaxi. As Uber's rivals in several of its biggest markets worldwide, the four announced they would connect their platforms, so that a user arriving in a new region could use a partner's service directly.

They formed a loose anti-Uber alliance.

Investment is what made this cross-border alliance possible. In May 2015, Didi put $100 million into a $530 million round for Lyft. It also took part in rounds for GrabTaxi and Ola.

Didi's overseas investments and partnerships

2015.5

Didi takes part in a new funding round for Lyft

2015.8

Didi takes part in Grab's $350 million Series E round

2015.9

Didi invests in Indian ride-hailing service Ola

2015.12

Didi, Grab, Lyft and Ola form a cross-border ride-hailing alliance

2016.4

Didi launches its overseas version, integrating Lyft's service

2016.8

Didi takes part in Grab's $750 million funding round

None of these companies is cheap, though, so Didi's stake in each is very small — nowhere near the share Uber held in Didi.

The alliance has also shifted with Didi's purchase of Uber China. Will Lyft, Uber's biggest rival in North America, keep sharing products, services and data with Didi? Uber, after all, cut the data link between its China staff and headquarters the moment the deal was announced. And most of Didi's investments in these companies were strategic stakes rather than large infusions of cash.

Asked in a Bloomberg TV interview a few days ago about the awkward overlap between partner and competitor, Lyft's CEO looked uncomfortable and said the details were still being worked out with Didi, so he wouldn't comment for now.

Where Should Didi Go?

Clay Shirky — author of the social-media bestseller Cognitive Surplus and an NYU professor who has lately been watching Chinese tech companies from China — gave us two ways to think about this.

"In practice, a Chinese company's global strategy will be a regional strategy first, the way Xiaomi has done better in India than in Brazil. It will be easier for a Chinese ride-hailing company to expand into Malaysia and the Philippines than to start in Nigeria or Argentina," he said.

The other direction is countries and regions that draw enough Chinese tourists. A startup there doesn't have to begin by winning local users; capturing what Chinese tourists use abroad already gives it some scale.

By that measure, Southeast Asia, Japan and South Korea all fit. Asian markets are close to home, relatively familiar, near enough in time zone to coordinate work, and easier to manage — much as Alipay and WeChat Pay expanded here first and are furthest along here.

The Easy Overseas Markets Are Already Crowded

Over the past two years, Uber has tried every market it could.

When it formally entered Beijing in July 2014, Travis Kalanick announced at a Sanlitun launch event: "Beijing is our 100th city."

Over the next two years, while fighting Didi, Uber moved into 400 cities around the world.

Uber's staggering valuation also helped its rivals attract enormous amounts of capital — every one of them hoping to become the next Didi.

In the hottest markets, Uber's rivals carry high valuations too.

Grab, Ola and Lyft — the larger members of the earlier "anti-Uber alliance" with Didi — sit in the markets where Uber has invested most: North America, India and Southeast Asia. The subsidy war China went through is being fought in those places as well, with investors following Uber's example and hoping to raise a Didi of their own in each one.

None of these is the weak market Liu described. With the money Didi has today, it would struggle to expand across several global markets, rebuild its China business and buy these companies all at once — unless Uber crushes them outright, in which case there may be no need to buy them.

These companies may not want to sell, either. At the same WSJD conference where Liu talked about going abroad, Lyft co-founder John Zimmer denied earlier reports that the company was looking for a buyer, and said Lyft's order volume had doubled over the past year.

But the Untapped Markets Aren't Smaller Than China

Ride-hailing is a business built on urban population; only big cities make a market large enough.

We counted the cities outside mainland China with populations above 2 million.

Outside mainland China, 1.231 billion people live in 215 cities of more than 2 million:

Of those, the United States, India and Southeast Asia already have local companies valued at $3 billion or more, and Uber is very active in each — India has become its largest market by order volume outside the United States.

That leaves 810 million people in 152 cities of more than 2 million. Bigger than the Chinese market already.

Apart from a few war-torn or extremely poor places, most of these markets are underdeveloped because of regulation.

Uber has been in Western Europe for years, for instance, but it is tightly restricted in most cities there. Nor has a single local premier-car startup there reached a $1 billion valuation.

Japan, Canada, South Korea and South America look much the same.

Japan bars non-professional drivers from carrying passengers, so for now Uber offers only phone-hailed taxis and food delivery there while it waits for the policy to mature.

All these markets will open. Premier cars plus carpooling run far more efficiently than taxis — even taxis hailed by phone. As one country's cities after another find their transport working better under the new model, other regions will change sooner or later.

North America shows this clearly. In 2012, Uber couldn't even get into New York, and its CEO said publicly at an event that corruption was the main obstacle to the technology spreading. Today New York City has adjusted its policy, and Uber has pushed the price of carpooling there down to a fraction of a traditional cab ride.

Quebec, to the north in Canada, is the same story: shut out in 2012, and now cleared to run a pilot.

Presidents, mayors and transport ministers change every few years; policy does not stay fixed forever. When these markets open, Uber will hold little first-mover advantage over Didi, and in theory the two companies will be back at the same starting line.

If It Really Does Go Overseas, This Could Be the Biggest Challenge in Didi's History

If Didi has to go abroad, it still has a lot to learn.

Uber today is valued at $68 billion and isn't profitable. That makes it easy to think all it does is burn money.

But the company, founded in 2009, didn't raise its first investment above $100 million until August 2013.

By then Uber had already spread to dozens of cities worldwide, probing each new one cheaply with the three-person teams it became known for: when Uber moved fast into a new city, the market was usually opened by three people — a city general manager, an operations manager who handled drivers, and a specialist for everything else.

The 500 cities Uber operates in today are spread over every inhabited continent, communicating across more than 20 time zones. Its engineering teams in Europe and the United States have spent years learning to build together.

Behind the capital and the loud CEO, Uber has become an extremely efficient multinational.

Didi, for its part, does not lack the ability to run a big company.

Professor Shirky noted that because the Chinese market is so large, Chinese companies reach enormous scale and learn how to operate a huge organization before they ever need to go abroad.

Didi's size isn't only headcount. By figures Liu disclosed this week, its system completes 20 million orders a day — taking the passenger's request, assigning the order, planning the route, navigating, collecting and paying out.

But the 400 cities Didi covers are all still in the UTC+8 time zone. It has no experience recruiting overseas teams to open markets, and its payment system can't even handle payment methods used outside China — an easy problem to solve, and one of countless problems it hasn't solved yet.

Its own operating efficiency is unresolved too. Behind the Didi app sits a heavily duplicated business-unit structure: building a new feature sometimes means building it again in every department (premier car and express car, for instance, are two of them).

Didi has just hired a group of people to tidy up that internal structure and make the company more efficient. It brought in Yu Jun, known as "the father of Baidu Tieba," as a product adviser; took Alibaba Cloud chief scientist Zhang Wensong from Alibaba to run its infrastructure platform and engineering; and recruited Silicon Valley security expert Gong Fengmin and cybersecurity expert Bu Zheng.

Roelof Botha, the well-known Silicon Valley investor and early backer of YouTube and Instagram, remarked recently that there is so much money in the market that many startups never have to face the final question.

"With that much money around, it's as if the exam has been postponed. You don't have to study any harder — you just cram later," he said.

Policy shifts at home, expansion abroad. What kind of company Didi becomes will depend on how it does on these "exams."

Cover image from CNBC, Visual Hunt