30 years of Internet recruitment: a war for efficiency

Internet companies are not only competing fiercely in commercial competition, but also in the competition for Internet talents. So are these companies just competing for talents?

At the end of 2021, LinkedIn China President Lu Jian had a bizarre day: a headhunter on LinkedIn wanted to recruit him to work as a backend developer. His experience quickly became a hot topic, along with the previous blunders of Meituan's Wang Xing being invited to become Wanda e-commerce CEO and Cheetah founder Fu Sheng being recruited to be an Android engineer.

The onlookers marveled at the "magical plot" and felt deep empathy at the same time - it turned out that they were not the only ones who were targeted by the mass recruitment information, but the big guys would also be accidentally hurt.

For HR and headhunters, “casting a wide net” is actually a basic operation. After all, in recruitment, inefficient matching “tortures” both job seekers and recruiters equally:

For workers, the most popular positions usually belong to those at the top of the pyramid; ordinary people are more likely to receive no response. The same is true for recruiters: large companies are overwhelmed with resumes, while small and medium-sized enterprises are often ignored.

As a result, top job seekers and recruiters will inevitably face harassment and a lot of screening work.

In fact, if we review the stories of "recruiting and finding jobs" since the birth of the Internet, we will find that the global online recruitment industry has been ups and downs for 30 years, and the service quality has always been less than satisfactory.

But the industry has never left the public and investors' sight. The rigid demand for human resources services determines that players can maintain a stable user and revenue scale. The TO B industry attribute determines that the industry cake is not small (the global market size will reach US$29 billion in 2023) and can also cross cycles.

However, improving matching efficiency has always been the driving force for industry changes and iterations, from overseas to domestic, from the Job Board (job advertisement) model represented by Monster and 51job, to the search model represented by Indeed, to the social model represented by LinkedIn, as well as new generation platforms such as ZipRecruiter and BOSS Direct, each generation of platforms that came later has a certain subversive significance .

The development history of this industry is essentially a "war history" centered on bilateral screening and matching efficiency.

1. “Matchmaking efficiency” is the fate of recruitment platforms

$5 million was a considerable amount of funding for Indeed, which just turned one year old in 2005, but it was not even a fraction of Monster's annual advertising budget.

As the world's first recruitment portal, Monster moved the "tofu block" recruitment ads from newspapers and magazines to the Internet, and within six years of its establishment, it occupied 60% of the online recruitment market in the United States. From 1999 to 2004, Monster firmly held the position of Super Bowl sponsor.

Therefore, when Indeed received the financing and prepared to make a big move, Monster's brand awareness was already outstanding. Not only was it not afraid of competition from Indeed, it also took the initiative to terminate its cooperation with the Super Bowl.

But the result is that the ant shook the elephant.

Monster's market share has since accelerated its decline, while Indeed, which rarely advertises, has seen a surge in the number of independent visitors, surpassing Monster in 2010 to become the largest recruitment website in the United States.

The interchange of their fates corresponds to the first "efficiency battle" among Internet recruitment platforms .

Monster's positions are actively posted by companies, and the number is relatively limited. It is troublesome for job seekers to browse jobs on different recruitment websites and company official websites.

Indeed is well versed in “borrowing strength to defeat strength” – in addition to the positions posted by companies themselves, the platform also “crawls” in job information from many other recruitment websites and corporate official websites. For example, job seekers can search for positions on Monster on Indeed.

This not only increases the traffic channels for Monster, but also increases the number of positions available for selection on the platform. It can even search in multiple dimensions such as industry, region and salary, thus maximizing the "one-stop" experience and making it more attractive to job seekers.

More job seekers will attract more recruiters to post jobs on the platform. The bilateral effect will be activated and Indeed will experience explosive growth.

However, although the Indeed model has achieved a "one-stop" job search, its essence is still the same active search as Monster, and the problem of inefficient matchmaking still exists.

After all, job seekers and recruiters often don’t know their real needs. If you ask them to search actively, they may not know whether their needs are reasonable.

This is like filling out college entrance examination applications. The pain point for students is not that they don’t know how to search for universities or majors, nor that they are dissatisfied with the number of applications, but that they don’t know which university or major is suitable for them.

Conventional thinking is to choose targets that are generally considered good. For example, it is not uncommon to see companies looking for talents who can build rockets at the price of screwing screws; job seekers are always looking for high positions and apply to famous companies according to their rankings.

In this case, leading companies and top talents may "pass by", not to mention most mid- and long-tail job seekers and small and medium-sized enterprises.

There is even a classic joke circulating in the recruitment industry: The HR of a large company receives hundreds of resumes every day. Seeing that the HR is overwhelmed, the supervisor throws half of the resumes into the trash can and counsels the stunned HR: "Luck is also part of ability."

Amid users’ dissatisfaction, the second “efficiency battle” among recruitment websites began .

LinkedIn, which was established at about the same time as Indeed, focuses on the social model, allowing both parties to manage their own LinkedIn accounts like managing their social circles. This not only makes information more transparent, but also allows for a deeper understanding of each other through personal connections, thus improving the efficiency of matchmaking.

And those who need the support of "connections" are usually the people at the top of the pyramid. The problem of mismatch of scarce talents that Monster and Indeed have not solved has been alleviated here on LinkedIn.

After the rise of LinkedIn, Monster's share of total job search traffic in the United States further dropped from 25.6% in 2011 to 9.9% in 2016, while LinkedIn entered an era of rapid growth in the number of users.

However, to date, the problem of matching the majority of suboptimal companies with suboptimal talent remains unresolved.

Before the previous wave was washed up on the beach, job seekers and recruiters could only choose to "endure" until the introduction of "technology" that was ignored by the industry - the overseas ZipRecruiter used mobile scenarios as an entry point and actively recommended candidates to employers through algorithms.

This caused a stir among online recruitment platforms, which were still mainly based on search and websites at the time, and also opened the curtain for the third "efficiency war" .

ZipRecruiter's statement is: "Traditional recruitment platforms focus on the top 1 million companies in the United States, and we focus on small and medium-sized enterprises."

Matchmaking is also different from the past: after job seekers have created their resumes, they can click "one-click application" to submit their resumes with the support of recommendation algorithms. The platform will "weigh" the job seekers' general level in the market and recommend companies and positions in a targeted manner.

On the enterprise side, it is also a similar "one-click publishing + intelligent recommendation" format.

In this way, the search mode's "strong first" rule of the game has been broken to a certain extent, which not only gives small and medium-sized enterprises and ordinary job seekers more opportunities to be seen and selected, but also improves the matching efficiency.

Back to the example of filling out college entrance examination applications, if Indeed allows candidates to search for available majors and schools themselves, LinkedIn allows friends and seniors to recommend majors and schools, and ZipRecruiter is the "Zhang Xuefeng" in the recruitment industry who understands the real needs of users better than they themselves.

The same is true for BOSS Direct, a domestic recommendation recruitment platform that also focuses on small and medium-sized enterprises. Since it was born in the mobile Internet era and added a direct chat function, its efficiency has been further improved. It has gradually become the leader in the domestic industry in the past 10 years.

However, the recommendation + direct chat model is still not the end point of the industry's evolution. After the AI ​​wave hits, the platform's data processing capabilities will be further improved, and the recommendations to both job seekers and job seekers will be more accurate. Recruitment at home and abroad has set off a new round of competition around AI.

The endless "technological involution" also means that the company's blood-making method must keep pace with the times.

2. The ability to serve “God” determines the upper limit of the platform

Contrary to what some people think, recruitment platforms actually rely on companies to pay for their services. That is, the more demand companies have for recruitment, the more money recruitment platforms make. On the contrary, if there are too many workers and too few jobs, recruitment platforms will have a hard time.

Generally speaking, companies are the "gods" of recruitment platforms. But at the end of 2022, Indeed did something that went against the grain: it imposed "disguised fines" on companies that read resumes but did not respond.

The incident originated from Indeed's newly launched "pay-per-performance" pricing model (PPA) - companies only pay the corresponding fees when job seekers apply for a certain position. If the application is not rejected within 72 hours, all applications will be charged.

In Indeed’s original conception, PPA was a “kill two birds with one stone” deal:

• Attract small and medium-sized enterprises: lower the payment threshold and avoid the embarrassment of paying but not receiving resumes;
• Increase the average order value of large enterprises: reduce the workload of resume screening, but also charge more;
• Platform side: increase monetization methods and improve monetization efficiency.

Prior to this, Indeed's monetization logic was similar to that of search platforms such as Google and Baidu, all of which revolved around traffic and exposure. Compared with Monster's model of building websites, cutting space, and charging companies "three fees" (advertising fees, resume viewing and downloading fees, and booth fees), the monetization efficiency of pay-per-exposure has reached a higher level.

However, the search model also determines that Indeed is difficult to get rid of its "dependence on large companies" in terms of customer groups. After all, only large companies care whether they are always at the top of the search results. Small and medium-sized enterprises are the main body of the market. For example, in the United States, 98% of companies are small and medium-sized enterprises with less than 50 employees, and large and medium-sized enterprises with more than 1,000 employees account for only 0.15%.

This is obviously not a good thing for Indeed, which needs to support a family. Indeed's parent company Recruit currently has a market value of US$88 billion. Based on the proportion of revenue contributed, the valuation of the part where Indeed is located is around US$26 billion, carrying the banner of growth.

PPA is Indeed's hope, but after the product was launched, large companies with overflowing mailboxes not only had to work harder, but also had to pay "fines" if they couldn't finish the work. Small and medium-sized enterprises did not receive as many resumes, but had fewer recruiters and could not confirm them in time, so they eventually had to bear a heavier recruitment burden.

Under pressure from all sides, Indeed had to suspend its PPA service.

Compared to Indeed, LinkedIn, a social model company, has evolved in its ability to make money, but it is also more deeply trapped in the "three-acre land" of major customers.

As early as 2005, LinkedIn launched a premium subscription service and started a "network building" business: ordinary users can only see their mutual friends, and "people you may know" are hidden, while paying users can see "friends of friends" and even "friends of friends of friends." (Liepin also has a similar monetization logic)

Under the temptation of paying for connections, by 2007, LinkedIn's premium subscription service revenue accounted for 53% of its total revenue.

However, as mentioned above, "connections" are more of a game for white-collar workers, white-collar workers and leading companies. The majority of small and medium-sized enterprises rarely need them and cannot afford them.

LinkedIn has chosen to be a top-tier recruitment company, so it can only covet the small and medium-sized enterprise pie. As a recommendation-based recruitment platform for small and medium-sized enterprises, it is a natural gold digger in this market.

ZipRecruiter used to live a small but beautiful life based on higher matching efficiency, charging for job postings, exposure, matching and clicks.

However, the overall recruitment environment in the United States has been chilly in the past two years, and companies have been cutting their recruitment budgets. As a result, ZipRecruiter, which requires payment to use, has become the target of "cuts", and its revenue has fallen to a certain extent.

Boss Direct Hire in China also has a paid program, but free is still the main theme.

After all, given the national conditions and market conditions, BOSS Direct cannot implement pay-as-you-go like ZipRecruiter; and its nature of serving small and medium-sized enterprises also prevents it from selling search results like Indeed.

In this context, BOSS Direct adopted a pricing method that sells "scarcity".

When the recruitment demand is not strong, the platform will lower the price or even directly list the positions for free. An HR of a startup once shared his experience on how to build a basic team of 30 people without spending a penny on the platform.

When there is a strong demand from recruiters, the platform will charge for posting related positions. For example, this year many real estate agents have switched careers, and agencies generally have difficulty recruiting. The recruitment fee rate for real estate agents on BOSS Zhipin has reached 100%.

In other words, the payment mechanism of domestic recommendation platforms changes flexibly according to the supply and demand situation between job seekers and recruiters.

This also requires the platform to accumulate more users and positions and stimulate the bilateral network. Otherwise, if the balance tilts too much to one side, this set of "price guidance" will fail. For example, when ChatGPT was popular last year, AI prompt engineer positions were snapped up, but the market talent reserve was far from sufficient. No matter how many "talent recommendation" props companies bought, the platform could not produce more resumes.

At present, BOSS Direct has established a certain "buffer zone" by relying on blue-collar workers and sinking markets, and delivered a satisfactory report card when the global recruitment platform revenue was under pressure last year.

But the subscriptions to ZipRecruiter and LinkedIn, the promotions on Indeed, and the allocation of scarce resources by BOSS Direct, cannot be considered the end.

The long-term iteration direction of online recruitment services is still aimed at the certainty of recruitment, that is, as long as employers pay, there will be suitable people available.

Currently, recruitment platforms pay based on exposure and performance, and in the future, payment based on results may appear.

Recruit disclosed a set of data in its financial report for fiscal year 2023: In 2023, the output value of global online recruitment services was US$29 billion, while the global talent dispatch, headhunting, temporary employment and other services that can "deliver results" had a cumulative output value of approximately US$224 billion.

This means that online recruitment has nearly 10 times the business potential, and these 10 times opportunities come from the grasp of "certainty".

3. Summary

In the "gold rush", those who dig for gold do not necessarily make money, only those who sell shovels make money.

The recruitment industry is a "shovel-selling" business - the rigid demand for employment determines that no matter whether the overall environment is prosperous or not, it is difficult for companies in the industry to go hungry.

However, Internet recruitment has experienced ups and downs over the past 30 years, and each generation has become more efficient than the previous one, but it has not produced giant companies like e-commerce, food delivery, and taxi-hailing platforms. This is because the business of "people looking for people" is far more complicated than "goods looking for people" or "cars looking for people."

The industry still needs to continue to revolutionize itself, and the changes in matching efficiency and business models also indicate that this may be a sleepless and endless change and reshuffle.

Author: Zhou Xiao and Zhang Ranran; Data support: Insight Data Research Institute

Source: WeChat public account "Look and Feel"

This article is authorized by @表外外里 to be published on Operation Party. Any reproduction without permission is prohibited.

The title image is from Unsplash, based on the CC0 agreement