Online recruiters take searches deeper with job-matching models

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

Job websites are morphing into job-matching platforms by using their members’ data to help give candidates and recruiters greater control of their searches.

US-based job search platform Indeed counts 350mn unique visitors a month and LinkedIn, the world’s largest professional network, reports that 62mn of its more than 950mn registered users visit its jobs pages each week.

These sites dominate the online recruitment market and they are now applying artificial intelligence to examine the wealth of data they have on their users in a bid to improve the search experience for candidates.

LinkedIn is deploying AI to better understand the intent behind each user’s search query.

“A lot more of what we’re doing is what I would call ‘intent data behind the profile’,” says Hari Srinivasan, vice-president of product at Linked­In. Here, the platform tries to predict which companies or jobs candidates might be interested in, other than those they overtly search for.

Indeed is using its members’ data in a similar way. “Sure, we extract structured information from résumés and job descriptions,” says Raj Mukherjee, executive vice-president for employer at Indeed, “but jobseekers and employers share so much more with us about their skills and preferences. We get intent and behavioural signals, data from two-way messages, and 12mn interviews.”

By identifying a user’s intent and better matching their skills to roles, the platform aims to display vacancies that jobseekers might otherwise have missed.

Sam Franklin, chief executive and co-founder of London-based tech recruitment company Otta, says its job-matching model gives candidates even greater control.

As well as their CVs, jobseekers can input their skills and preferences, such as salary expectations, working patterns, and values. These data points are then used to match candidates to a curated selection of jobs, which Franklin says will be more relevant. The model is “supportive of jobseekers, supportive of topics like diversity and inclusion”, he argues.

Franklin says candidates’ perception of an employer’s brand is increasingly important. So, for each job posting, he or co-founder and chief operating officer Theo Margolius write “Otta’s take” — an overview of the hiring employer. Details listed include notes on employer response times, employee endorsements, funding rounds, and diversity and inclusion initiatives.

Mark Chaffey, chief executive of hackajob, another tech-focused UK jobs aggregator, agrees that corporate reputations are increasingly important to candidates.

Jobseekers who sign up to hackajob can follow companies, like they would on LinkedIn, and view content they post, but it is the companies that approach potential candidates about a job.

“It’s a way to change the engagement dynamic,” Chaffey says. “A candidate can show interest in an organisation, but not necessarily apply for a job. It’s more the idea that you [as an employer] can then share events, podcasts, and content to ‘warm’ that candidate.”

Hackajob’s candidates, who indicate preferences in areas such as salary and sector, are screened before they are made visible to employers. This means companies see only suitable candidates, says Chaffey, while jobseekers are only approached by companies that meet their expectations in “tech stack, location, hybrid policy”. He adds that 90 per cent of candidates respond to recruiters’ initial outreach within seven days.

Unlike platforms such as LinkedIn, hackajob charges clients a subscription. Chaffey says its “level of candidate engagement” enables clients to achieve significant cost savings compared with the fees charged by traditional recruitment agencies.

Alongside self-search, LinkedIn, Indeed and Otta also allow recruiters to contact candidates who are not actively job hunting. In addition, LinkedIn is adding an AI tool that will allow recruiters to conduct searches using ‘natural language’ prompts, such as “I want a London-based data analyst”.

It’s much more intuitive,” says Srinivasan. “We can start suggesting companies or things you’ve never heard of.”

But skills are an increasingly important element of focus, for candidates and recruiters alike. LinkedIn members added more than 500mn skills to their profiles over the past 12 months, Srinivasan says.

“People are getting hired a lot more based on their skills,” he says. “We have had hundreds of millions coming in and adding proof to their skills [to their profile]. And we’re seeing about 40 per cent of recruiters now coming in and . . . saying, these are the skills I need.”