Startups Are Out — Why Has Co-working Become a Big-Company Business?

That seems to hold whether you run the space or rent a desk in it.

Small startups of five or six people, setting out to go from zero to one, may no longer be the customers this young "co-working" industry wants.

Read any coverage of co-working from five years ago or earlier and the emphasis is still on innovation and entrepreneurship; the tenants named are usually early-stage startup teams and freelancers.

In 2010, when P2 (People Squared) opened its first co-working project in a lane in Shanghai's Jing'an District, it billed itself as a place for "entrepreneurs talking to entrepreneurs." By 2013 a new crop of spaces aimed at the "maker" crowd had appeared, Shanghai's XinCheJian among them. WeWork, in the same years, was just preparing to expand from New York to Boston, and the companies moving in were mostly small tech startups and design firms.

When co-working first turned up as a fashionable idea, it was almost always about starting a company.

When Jonathan Seliger became CEO of naked Hub last August, an order for five desks was enough to set the team, then less than a year old, cheering. Now what its members talk about is orders for several hundred desks — "those startups that just raised a Series A." WeWork founder Adam Neumann, by contrast, was still sounding careful about big companies moving in two years ago: "We've held back, tried not to overdo it," he said. Today "enterprise members" — 20% of WeWork's members worldwide and 30% of its revenue — are one of the things it promotes hardest.

Even when SOHO China talks about startups, it means something different. It launched its co-working product, SOHO 3Q, in February 2015. The startup tenants Pan Shiyi cited from the beginning were companies that had already reached scale and were expanding fast: the standard example used to be LeEco, which once leased nearly 1,000 seats at SOHO 3Q. Xiaohongshu, another name he likes to cite, once took half a floor at Fuxing SOHO 3Q; now it fills the whole Shanghai Fuxing SOHO 3Q building.

Operators are of course glad to see their customers get bigger. Serving one company that takes 200 desks is plainly more efficient than serving 100 companies that take two desks each.

And big customers mean larger orders on longer leases.

1.

Beyond the few hundred desks a startup headquarters takes, Pan Shiyi also wants SOHO 3Q to sink into smaller cities alongside companies like Zuoyehezi — an education app for primary-school students that has closed a 100-million-yuan Series B — as they push their sales channels down there.

"No need to worry about going to second-tier cities." As Pan Shiyi sees it, many of SOHO 3Q's customers are platform companies like Zuoyehezi, and every one of them needs branch offices in provincial capitals. He told QDaily that Zuoyehezi had put it to him directly: if SOHO 3Q opened in other cities, all they would have to do is rent SOHO 3Q desks in each one.

P2 founder Zheng Jianling makes the same point from the other end. When a tenant grows and needs to enter several cities, operators are too small to follow it — one of the industry's challenges, and also its opportunity. Seven years after it was founded, P2 is in only five Chinese cities, running 25 locations.

But that isn't the only reason Pan Shiyi likes the second-tier city market.

During the interview he ran the numbers out loud: Changsha has more than 7 million people, and at least a third of them — over 2 million — are working. Yet the city's office buildings come to only a little more than 2 million square meters. At 10 square meters a person, that puts just 200,000 people in offices. "Where did everyone else go?" He had been to Changsha to look around not long before we spoke.

Office vacancy rates in second-tier cities are in fact not low. 2017 China Real Estate Market Outlook, published early this year by the property consultancy CBRE, predicted that among second-tier cities only Nanjing, Hangzhou, Ningbo and Dalian would stay below 20% vacancy this year. According to its Global Shopping Centre Development report, China built nearly half of the world's new shopping-center floor space last year, while retail vacancy in first- and second-tier cities keeps climbing.

To Pan Shiyi, all this slow-moving commercial property already standing in city centers has become a resource that can be "converted" into co-working space.

After two and a half years of operation, SOHO 3Q announced this August that its product and management model had matured and that it was formally beginning a nationwide expansion by leasing space from other developers. In late September it held a press conference for the purpose, announcing three winning bids in Shanghai, Nanjing and Hangzhou and partnerships with Xindi Group and Shanghai Land Group — only the first batch of partners.

Pan Shiyi now spends all his time touring second-tier cities. "This is the one thing I'm busy with right now," he said. In one month he covered eight cities including Changsha, "looking at seven or eight projects in each one, sometimes more than a dozen."

One of his findings is that conventional office buildings cannot meet the needs of small and mid-sized businesses. "The people who own office buildings always say, 'I want to serve large and mid-sized state-owned enterprises, I want to serve financial institutions.'" He thinks SOHO 3Q can use the idle property in second-tier cities to fix this mismatch between supply and demand.

The other thing that struck him is how the gap between office workers in first- and second-tier cities is closing: the topics young people discuss, the apps they use, even the way they live, are converging. The internet has connected a great deal, and for the co-working business that is good news.

2.

Location, location, location.

Real estate's golden rule, seemingly the only standard that has ever counted, is being copied in co-working.

WeWork, now open in 17 countries, still talks about Manhattan a lot. Aaron Fritsch, WeWork's head of product systems and operations, said in an interview last month: "I don't think there's anywhere in Manhattan where you could open a WeWork and not draw a crowd."

Since the first WeWork opened in Manhattan's SoHo in 2010, Manhattan has come to hold 40 of them. The first WeWork with a single tenant — IBM — opened in Manhattan too (IBM was also WeWork's first traditional big-company client). And UrWork, the Chinese operator WeWork sued over the resemblance of its name, plans to open its first New York location in Manhattan as well.

naked Hub, which bought the Singaporean co-working brand JustCo this July, says it is still hunting for acquisitions, and that the first thing it looks for is whether the target holds property in good locations. CEO Jonathan Seliger told QDaily: "Location matters most. I wouldn't buy a company in Zhangjiang (Shanghai) — I'd buy one in Jing'an or Xuhui."

SOHO 3Q, whose bidding documents mark out preferred and second-choice zones for each city, likewise requires sites in "reasonably busy areas." Pan Shiyi's more exact version is "a five- or six-minute walk from a subway station."

That is because all of them treat conventional office buildings as the competition.

WeWork founder Neumann said in an interview three years ago: "We're not competing with other co-working brands. We're competing with office buildings."

Seliger, in a fifth-floor meeting room at naked Hub's office on Loushanguan Road in Shanghai's Changning District, told QDaily: "Look at the buildings out the window — those are our competitors. The companies you mentioned, the other co-working companies, are not."

This shows up most clearly in the companies that turned from incubators into co-working spaces.

KrSpace, founded in April 2014, began as an incubator: it screened the companies it let in but asked them for no fees and no equity, and it mostly picked sites in areas where the government paid subsidies — KrSpace in Suzhou, for instance, sits in the Suzhou Industrial Park, far from the city center.

36Kr founder Liu Chengcheng told QDaily that once KrSpace began turning into a WeWork-style co-working business at the end of 2015, the first thing it looked for in a site became "a central area with good transport," whatever the land cost.

Clearly, to take share that used to belong to office buildings you need an advantage in location, and the high-tech parks local governments lay out usually sit far from the center. Those parks also come with government subsidies; co-working spaces get none. In its second-tier city push, SOHO 3Q steers clear of government-planned high-tech parks, because "it isn't a fair fight."

3.

Only bigger companies — bigger capital — can pay for good locations and the cost of expanding on a large scale at the same time.

This August WeWork raised $4.4 billion from SoftBank, $500 million of it earmarked for the China market. The same month UrWork raised 1.2 billion yuan. KrSpace, which raised 100 million yuan last month, says it wants to raise and spend 4.5 billion yuan from here on, rather than continue "playing small" as it used to.

SOHO 3Q, for its part, has taken no outside money yet. In June SOHO China sold Hongkou SOHO for 3.573 billion yuan; add the two Shanghai properties it is preparing to sell and the cash it will have raised from selling buildings over four years comes to roughly 34 billion yuan.

In the interview Pan Shiyi again insisted this is not capital flight, and took the chance to say he is confident about the Chinese market. As he sees it, putting the proceeds from selling buildings into an asset-light industry is worth doing — "there's no funding pressure even if we open in another 8 or 16 cities." Once he is satisfied the model works and is ready to expand into every city in the country, SOHO 3Q will open itself to outside investors.

Co-working has gradually become a game only big companies are qualified to play.

It isn't only the huge sums needed to cover the cost of expanding; it is also the cost advantage that comes with scale. WeWork chairman Artie Minson once explained why they don't care about small companies: "The bigger you get, the cheaper and more efficient it is to build [offices], and the more financial flexibility you have when you negotiate a deal."

Pan Shiyi says that when SOHO 3Q started in 2015, fitting out a single desk cost about 30,000 yuan. Now that it has 16,695 desks, cheaper materials and better design have brought the average down to about 15,000 yuan a desk. The three 3Q projects he calls failures — Donghai Plaza, Danling SOHO and Zhongguancun SOHO — have one thing in common: they are very small, around 200 desks each, and their average costs run too high.

Meanwhile most operators are now eager to talk about things other than desks, and many of those services only a bigger company can afford to provide.

naked Group, which started in resorts, is the clearest case. naked Hub calls itself a lifestyle brand, and what it offers is not confined to work: restaurants on site or discounts at restaurants nearby, yoga and meditation classes, roof terraces that can hold events, and more.

Seliger says many people on naked Hub's service team come from five-star hotels or retail. They have even launched an executive-assistant service called "naked Angel," which will buy a member coffee, book a hotel or find a lawyer.

naked Hub, which has 20,000 seats today, also keeps a design team and a technology team of 50 people each. Both take feedback straight from members and staff and use it to improve the design, and having designers in-house lets naked Hub finish drawings fast enough to keep the expansion plan on schedule.

Its membership numbers also give naked Hub bargaining power. Plenty of co-working companies sell their members tax, HR and insurance services and take a commission. The more members a brand has, obviously, the lower the price it can negotiate, and the more appealing it looks to those members.

The Service Store WeWork launched in the U.S. this April sells more than 100 software services — Office 360, Slack, Avis car rental and others — to its members at a discount as well.

4.

But just as it takes a large property company to run several plots of land well, experience in the trade, staff to deliver the service, and the ability to attract talent are all needed to run co-working well.

Co-working has already shown it can pull in senior people. In 2015 Arthur Minson, then Time Warner's finance chief, decided to become WeWork's CFO. Seliger, naked Hub's CEO since last August, was global CEO of the Italian jeans brand Miss Sixty and China CEO of Coach.

Their ambitious expansion plans in China also depend on solving the talent problem.

According to a Bloomberg report in May, WeWork has close to 2,400 employees. Its website currently lists more than 500 open positions worldwide, 37 of them in China. KrSpace, and naked Hub — which is "hiring like crazy" to keep up with its expansion — each list more than 20.

5.

Every co-working brand has its own picture of where the industry, and its own company, is going.

Pan Shiyi says "this is an internet business built on real estate." Around 80% of SOHO 3Q's customers are internet companies, yet the way SOHO 3Q operates mostly copies SOHO China's property experience. What it is working on now is standardizing the product and the leasing process — the most urgent job at the moment.

naked Hub, which comes out of the hotel business, has transplanted the service model of its naked Stables resort. Providing "hotel-level service" is what Seliger calls the biggest difference between naked Hub and everyone else; he thinks co-working will end up a service industry. So although rent is still naked Hub's largest source of revenue, what excites him more is the fast growth in income from renting out event space and the commissions from selling more services, including software (again built on what it learned running hotel management systems).

And after landing IBM as a major client, WeWork has come to look more and more like a property consulting and management firm. WeWork China told QDaily that the company has a design and development team of three to four hundred people worldwide who can find a building, design it, build it out and run it for a company that needs one; a client can pick any one of those steps or several, and the service will be available in China soon.

As every co-working brand chases more big companies, its competition with office buildings will grow more direct — and the difference between the two will keep narrowing.

"Big-company clients mean filling the space faster, filling the space faster means profit faster and new locations faster. This really is all about taking market share."

These companies all have large goals.

WeWork plans to enter three to five new cities on the mainland next year, and will keep expanding in Beijing and Shanghai as well;

naked Hub will go from 5 locations today to at least 200 within two to three years;

KrSpace plans to grow from 100,000 square meters today to 1.5 million square meters by 2019;

UrWork plans to reach 180 locations within three to five years, with more than 100,000 desks.

SOHO 3Q hopes to raise its desk count from 17,000 to 170,000 within three to five years.

This is what the co-working business has turned into: property, hotels, office leasing, property consulting? It looks like a bit of each.

If you had to sum the industry up, these might be its traits (and they may apply to any industry):

• Swallowing the whole chain from upstream to downstream in one gulp no longer works. Every company is now defining who it sells to, and none expects to win every kind of client from the cash-strapped startup to the Fortune 500.

• Location, location, location.

• The big get bigger; the winner takes all.

• The 80/20 rule. The customer base still ends up dominated by big companies. Small ones can't produce economies of scale, and the most profitable customers are still the big ones.

• No successful large property company was ever built on "I already hold land," so the idea that "I have old factories and old buildings, so I have an edge in co-working" deserves a large question mark. This is not simply a business of subletting.

• Nor, along the vertical chain, can any single company take all of it — development, property management, leasing, service. Nothing like Wanda Plaza, a big company combining commercial and residential property, has appeared here yet.

Given all that, and even though a small startup may generate very thin marginal returns, the co-working spaces all say they won't give up that kind of customer entirely. Variety, and a complicated ecosystem, is what they want to build.

At naked Hub, where multinationals dominate the client list at 40%, startups still account for 15%. SOHO 3Q, where IT and tech companies make up close to half the tenants, would like a few third-party service firms. And even KrSpace, which spun out of 36Kr and wears a stronger "startup" label, keeps internet companies to no more than 50% of its customers.

As for the "community" they all used to talk about, none of them has forgotten it — but each defines it differently.

Cover image: naked Hub, WeWork, adsttc