Job-hopping season is here, but the recruitment sites aren't doing so well this year

The way recruitment sites are remaking themselves is a microcosm of the entire internet industry

March and April are job-hopping season. The new year has passed, the year-end bonus is in hand, and every office worker who is unhappy with things starts thinking about a move.

Every year at this point, recruitment sites like Lagou, Liepin and Zhaopin plaster subway and bus stations with giant ads, timed to the online and offline "Promotion Season" campaigns they have spent months preparing.

But this year the industry has mostly been getting bad news.

Last Friday, US-listed Zhaopin received its third take-private offer. Unlike the first two, this one may actually go through. A special committee inside Zhaopin is in talks with a buyer group made up of its largest shareholder, SEEK International, along with Hillhouse Capital and FountainVest Partners; the group would buy every outstanding Zhaopin share it does not already hold.

The outlook is not good. This past Tuesday, Zhaopin published its results for the last three months of 2016: a net loss of 5.2 million yuan for the quarter, against a net profit of 61.7 million yuan a year earlier. The loss means that even if the take-private goes smoothly, Zhaopin cannot immediately file to list on the A-share market.

The slide did not come out of nowhere. Since the third quarter of 2015, even in the quarters when Zhaopin's net profit grew, it grew only about 10-20% year over year. Earlier the figure had been 30-50%.

Zhaopin's run on the US market has in fact always been lukewarm. The share price has gone from its 2014 IPO price of $15.34 to $16.92 today, and its all-time high was only $18.50.

Zhaopin is already one of the largest online recruitment companies in China. According to Analysys, as of June 2016 it ranked second with 27.5% of the market. By mobile app activity, it ranked first.

51job, whose share is a little higher still, is sliding too. Its third-quarter 2016 results showed net profit down 35% from a year earlier.

Smaller companies have a harder time staying alive — take Zhoubotong Recruitment, which has already shut down. A former executive who left before the company closed told QDaily that Zhoubotong stopped operating in the second half of 2016; nobody switched off the servers, so the site still loads. You can still find a handful of fresh job postings on it, but Zhoubotong's own hiring ads stopped eight months ago.

Lagou, which recruits for internet companies, has lately been talking about extending its business too. Last year it launched two products, Dakun and Lagou Cloud HR. The first connects pay-as-you-go talent with the companies that need them; the second is a cloud-based HR management system for corporate clients.

Companies are always hiring and someone is always looking for work, so recruitment ought to be as basic a business as food, clothing, housing and transport. China produced 7.65 million college graduates in 2016, 160,000 more than the year before, and in theory nearly all of them use recruitment sites — before you even count the people already employed and thinking about a move.

Two or three years ago this industry was a lively place.

2014 was the busiest year for recruitment sites, and the most feverish year for startup investing

The big change in looking for work online came in 2014. Zhoubotong Recruitment was part of it.

In 2007, Feng Tao, then an engineer at IBM, started a "Zhoubotong Recruitment" group on Douban. As in any Douban group, members traded information and swapped notes on their own.

But it gathered members far faster than anyone expected, reaching tens of thousands within two years. HR staff and headhunters began posting openings and approaching candidates there regularly.

In 2011, Feng Tao turned the group into a business: he found co-founders, registered a company, and launched the Zhoubotong Recruitment website. In 2014 he raised a 28 million yuan Series A from NetEase, which gave Zhoubotong plenty of money to burn.

The good times lasted about a year. By the end of 2015, reports said Zhoubotong had laid off a third of its staff after a fundraising plan fell through. Feng Tao replied that "the cash on hand is enough to last at least a year and a half."

The final bad news came sooner than that. Zhoubotong stopped operating in May 2016.

Even at its best, Zhoubotong never built a decent product. The whole site read like a Douban group moved over wholesale. It egged employers on to write clickbait posts, so you got headlines like "Unlimited snacks! Company trips every year!"

Its mobile app lifted the logic of the dating app Tinder outright: job seekers and companies swiped left and right to pick each other. But hiring and job hunting are not as simple as judging a face.

Like Zhoubotong, Lagou and Liepin — the two biggest "vertical" recruitment sites — also caught on in 2014.

Search "recruitment" in the database of the startup-investment service IT Juzi and more than 600 companies come up.

Grouped by how they connect HR staff with job seekers, they fall into roughly four types:

• Algorithmic matching that recommends suitable jobs to candidates: Lagou
• Headhunting moved online: Liepin
• Talent auctions, meaning screening resumes and pushing them to employers: 100Offer, Nextoffer
• Job hunting through social networks: Boss Zhipin, Neitui

They caught the moment when internet companies had the most money. Investment in the first half of 2015 alone matched the total for all of 2014. There were 3,932 funding deals in 2015; five years earlier there had been 302.

"Demand from the internet industry was huge — almost everyone needed to hire," said Han Bing, founder of Nextoffer, who started Nextoffer around this time.

This was also when online recruitment sites like Lagou and Liepin began to grow. All of them raised money between 2014 and 2015, and as other startups in the industry needed to hire more and more people, the recruiters' business grew with them.

The new services changed the product, but they cover only one part of hiring

Recruitment sites are a product of the PC internet. 51job, Zhaopin, ChinaHR and 58.com/Ganji are all more than a decade old, and their core business is standing between job seekers and employers, charging a commission on information the two sides lack.

A company's HR staff can post openings there, buy ad space, and buy job seekers' resumes. That is where the money comes from. According to Zhaopin's filings, 90% of its revenue comes from online recruitment. The rest comes from campus recruitment, assessment services, headhunting and so on.

Even today you can find homepages that look like a newspaper's classified page, packed tight with banner ads for everything from hourly and blue-collar work to senior talent in internet and finance.

In the story that "vertical" recruitment sites like Lagou and Liepin tell the public, the traditional sites are riddled with old problems and run inefficiently.

"Traditional recruitment websites are sales-driven companies, and the huge volume of redundant information they generate makes them less and less efficient — nowhere near what emerging industries need," Feng Tao said just after raising money in 2014. "Splitting the market by industry, with the user at the center, is the direction online recruitment will take."

The founders of Lagou, Liepin, Boss Zhipin and 100Offer have said much the same thing. "Big data," "precise matching" and "vertical field" come up again and again.

• Lagou: hiring for internet jobs, built around "precise matching," with revenue mainly from value-added service fees charged to employers. It does not sell resumes.
• Liepin: mid- to senior-level positions, headhunting moved online, with revenue mainly from the fees headhunters and HR staff pay to view resumes, plus value-added fees charged to users, such as pinning a resume to the top.
• 100Offer: mostly internet product roles. It screens candidates and pushes them to employers, charging headhunters and HR 12% of a successful hire's annual salary.
• Boss Zhipin: hiring for internet jobs. At bottom it is an instant messaging tool that lets job seekers and HR talk directly inside the app, with revenue mainly from value-added fees charged to job seekers.

Judged by the result, these services solved a good part of the problem. The experience got more comfortable, at least for job seekers.

You no longer see the huge bulletin-board ads, for one thing. The interface is far cleaner, job recommendations come in an orderly list, and looking for work feels a little less like hunting for a needle in the ocean.

For another: plenty of employers would rather not state a salary range, and Lagou forces them to. Employers get "zombie resumes" that go nowhere, and both Lagou and Liepin use one product trick after another to push users to fill their resumes out properly.

Going from a crude traffic-and-advertising model to a business built on service and experience looks almost inevitable. Entrepreneurs in the industry like to cite one case as proof: Monster, the earliest and largest recruitment site in the US, lost 90% of its peak value by 2014, precisely because the vertical player Dice took its market share.

But that reversal has not happened in China. The sites with the most traffic still have the biggest shares. According to Analysys, in the second quarter of 2016, 51job held 29.1% of China's online recruitment market, Zhaopin 27.5% and 58.com/Ganji 25.4%. All of the newer recruitment sites put together held 18.0%.

The reason may be that improving the product did not change the underlying logic of the online recruitment business.

"[Online recruitment] is a business model that is very hard to quantify, and it leans heavily on traffic," Han Bing said. What he had done before — "narrowing the chain, screening resumes, raising the matching rate" — was still only "moving the work of a headhunting firm online."

They may no longer sell ad space in bulk, but recruitment sites are still selling information, and that takes traffic and users. The more users submit resumes and fill in their details, the bigger the database a site has to match against. A larger user base also means more people to convert into paying customers for value-added services.

At bottom this is no different from Zhaopin, ChinaHR and the other traditional companies the newcomers say they are going to overthrow.

For every internet company, the price of traffic and the cost of acquiring a user keep climbing. Traditional companies like Zhaopin and ChinaHR, which made their names earlier and have more users, actually feel that pressure less than the newcomers do.

The problem employers most need solved is even further from being solved

The business models online recruitment uses to make money cannot keep up with what employers want.

"I'll try the recruitment sites first, as a way in. If I find someone online, great; if not, fine," a headhunter at a Guangzhou firm specializing in media told QDaily. In her daily work, only 30-40% of her "hunting targets" come straight from sites like Liepin, Lagou and Zhaopin.

Getting to a candidate is a well-practiced routine. This headhunter works out a company's staffing structure herself. A candidate who turns her down is not worthless to her: starting from that person, she can gradually talk her way to a picture of the company's business units and who sits in them. Or she asks for referrals to friends and peers — "nine times out of ten, someone else's recommendation is right."

An HR staffer who used to work at a well-known internet community company with several hundred employees told QDaily that he went through about 1,200 resumes a month from recruitment-site channels, counting both unsolicited applications and ones he pulled out of resume databases. Of those 1,200, "it would be pretty good if 10 ended up hired." Most months, "not one works out."

From referrals and headhunters he received only about 40 resumes a month, but three of those would typically be hired — "far, far more accurate."

For any headhunter or HR person, the goal is to hire someone at the lowest possible cost in conversation. That means the work depends on people, and on living with a lot of uncertainty.

The newer sites all talk about matching and screening talent, but their basic function still stops at connecting a person to a company. Hiring takes a long time. Who can guarantee the two sides will hit it off and sign happily?

That creates another bind. People do not use recruitment sites often, most job seekers do not post their resume on only one site, and the same resumes turn up across different sites.

"Recruitment sites aren't the best option — they're really advertising sites," Han Bing said. When a company pays a service provider, it weighs the odds of "actually being able to hire someone": "Rather than spend a little money on low odds, you're better off spending more on high odds." Referrals, introductions through people you know, and hunting on vertical communities like GitHub and LinkedIn are all "higher-odds" routes.

"It is very hard for a new online recruitment company to pull off the kind of disruption Didi pulled off with taxis."

The founders are not blind to the limit. Lagou CEO Ma Delong told QDaily that he wants "every resume that goes out to get an answer," and that what Lagou can actually do still comes down to data and algorithms: "Does it help? Yes. Is it an earth-shaking change? I don't think so — not unless you put a chip inside a person's body, so I know what they're thinking."

Once the internet startup boom passed, the problems arrived

Recruitment may be the industry that reflects the state of the economy most directly.

China's GDP grew at exactly 6.7% for three quarters of 2016, the lowest rate in 26 years.

Accordingly, the China Employer Demand and White-Collar Talent Supply Report that Zhaopin published at the end of 2016 found that both hiring demand and resume submissions fell nationwide — but resume volume fell far less than the number of openings. Openings dropped; job seekers stayed relatively active.

The change in the internet industry is easier to see. Think back over hiring and employment in the industry in the past year and a few headlines come to mind: Alibaba abruptly cutting the number of campus offers it made; Baidu halting experienced-hire recruiting; Tencent halting outsourced hiring.

LeEco hit a cash crunch. Didi lost most of its drivers under the new ride-hailing rules. Xiaomi's phone shipments kept falling. These companies, each valued at more than 10 billion yuan, had been expanding fast and hiring in bulk. Didi, for one, rode subsidies, expansion, mergers and dominance to become a genuinely large company in under five years: 6,000 employees and a $38 billion valuation.

Meanwhile the list of dead startups grew by another string of names: Kaola Banche, Mitao, Meiwei77, Qingnian Caijun.

Pay felt it directly. According to the venture firm IDG's 2016 China Pre-Unicorn Company Compensation Survey Report, a slack startup funding market and large-scale layoffs cut average income for internet business and marketing roles by 45% compared with 2015.

The whole internet industry is no longer as frantic as it was two years ago, and growth has slowed. "Lagou is still in a growth phase, we haven't hit a ceiling, we're still growing — but the rate isn't as extreme as it was. The industry itself is in a period of fluctuation," Ma Delong said. Mass layoffs sped up the movement of talent, which turned out to be an opportunity: Lagou's business volume is still rising. Its Internet Workplace White Paper says the number of people working in the internet industry rose 12.3% in 2016 from the year before, and pay rose too.

New businesses are brewing, but the old model has not changed

Since the hard problem cannot be solved for now, recruitment sites have started trying other things.

Han Bing has essentially stopped putting people behind Nextoffer; his new product is a cloud-based management system for HR staff called Daxiang HR.

Lagou launched "Lagou Cloud HR" at the end of last year, and internally the new business carries enormous weight. Ma Delong set the team up as a "business unit," on the same level as the original recruitment business, with its own engineers, sales staff and operations staff.

Every one of Lagou's corporate clients now lands straight in this new platform when they open the site's back end.

According to Ma Delong, "Cloud HR" exists to "help a person grow inside a company," not just to solve hiring and employment. It is an "extension and expansion" of the original recruitment business.

The system has modules for personnel management, recruitment management and more, letting HR and management handle everything from hiring and onboarding to promotion, pay and performance reviews in the cloud. Once an employee is hired, HR can send electronic documents through "Cloud HR" to finish the onboarding paperwork rather than carrying paper forms back and forth. Pay stubs can be approved and issued quickly too, replacing the printed ones.

This looks a long way from the original recruitment business, but the motive is not hard to read. Lagou wants its commercial return to come more directly — charge the employer, hold on to a doorway through which services can be sold, and ideally get small and mid-sized employers using its cloud management system all the time, instead of opening a recruitment site only when they need to hire.

But HR management does not bring a company revenue directly. Big companies usually have their own management tools already. At small companies, it is hard to turn into something anyone pays for. It exists mainly to solve a traffic problem.

Now that internet entrepreneurship has cooled, online recruitment will not find it easy to heat up again.

Cover image from Pixabay.