Artificial Intelligence & Recruiting Data

Terry Elliott Products & Technology

Significant changes are occurring in the sources of data that recruiters use to find candidates.

  • LinkedIn has been purchased by Microsoft.
  • Indeed was purchased by Recruit Holdings.
  • Monster is being purchased by Randstad.
  • CareerBuilder is being purchased by GTCR.

Although most people think that these database services will continue to be operated as independent businesses in order to keep the data current, it is also clear that Recruit and Randstad, plan to leverage the databases that they own in order to create a competitive advantage for their agency businesses. They plan to do this through applying advanced search and matching technology based on Artificial Intelligence (AI) and computational linguistics and these services are not expected to be offered to competitors.

There are always rumors. We have heard a rumor that Recruit has begun to limit usage of Indeed for one of the 5 largest staffing companies in the world. We also heard a rumor that Seek, the dominant job board in Australia, restricted access for an applicant tracking systems vendor that competes with JobAdder, a vendor in which Seek is an investor. Randstad has gone out of its way to assure users of the Monster job posting and resume database services will continue to be operated as a separate and independent entity.

Although access to big data can represent a competitive advantage for recruiters, the changes underway in the systems that are used to search and match candidates can offset the advantage of big data with more limited search and matching capabilities.

Google Jobs is an example of an AI application for searching and matching that could leapfrog the value of restricted access to candidate data. The system is currently being developed with the Google search engine, the CareerBuilder database and the Google Translate neutral networking system. It is scheduled for beta test release in the 3 rd quarter 0f 2017.

Google Jobs uses a cluster of descriptive terms to more accurately match candidates with positions, instead of the single terms (like “Java programmer”) that are primarily used in searches today. Jobs and candidates can be found by replicating the same scraping and aggregation technologies used to build Indeed, pairing them with much more accurate search and matching technology. Part of Monster’s attraction to Randstad is that Monster had also been adapting its matching technology to an in-house RPO business. IBM Kenexa is applying the Watson AI system for its RPO customers.

Recruit Holdings is also developing much more accurate search and matching technology. Recruit set up the Recruit Institute of Technology headed by an AI authority from Google Research in late 2015 to work with the Indeed data that they acquired in 2012. We suspect that their intent is to use this capability internally for their own competitive advantage. Recruit’s goal is to become the Number #1 staffing company in the world by 2020. They have a major hiring program underway and we believe that their plan is to use this data to move into higher value-added services.

The strategies that are emerging for the major agencies like Recruit and Randstad, as well as technology-based firms like IBM Kenexa, are to not so much to control the data but to create competitive advantage by developing advanced searching and matching technology for such higher value-added services and RPO, VMS and staffing.

It is easy to miss that these blockbuster acquisitions are not just about the data that recruiters require. Microsoft didn’t pay $27 billion for LinkedIn’s $175 million job posting business. Microsoft wants the “content” (the profile information) for 467 million professionals worldwide for their CRM (Customer Relationship Management) applications. Microsoft can advance the development of their searching and matching technology (which is based on Artificial Intelligence and computational linguistics) with this data, but the data isn’t the end game. Microsoft wants to offer a unique value that their CRM competitors cannot offer and Microsoft wants to offer a unique value that their cloud competitors cannot offer and they need this data to develop the AI value-added.

In order to build that value, according to conventional reasoning, Microsoft, Recruit, and Randstad have to keep that content current and this requires continuing to offer recruitment advertising services to corporate recruiters and agencies. This means offering both job postings and resume database searches. So what is the concern for recruiters? The concern is that the current state of the art for resume searching and matching is incredibly crude; the data will help these vendors to develop better search and matching technology but there is no reason to share these breakthroughs with competitors.

For example, Facebook is one of the largest employers of computation linguists and AI experts in the Silicon Valley and many recruiters now rely on Facebook more than LinkedIn. Also, Oracle (Taleo and PeopleSoft) has a major AI, business intelligence, and analytics platform program under development. Also, Salesforce’ acquisition of BeyondCore adds real teeth in their Einstein AI program.

Other competitors with advanced search and matching technology will follow, however, Google Jobs may be the most advanced with a beta test version scheduled for release in the 3 rd quarter of this year. We assume that Google Jobs with being delivered via an API, like IBM’s Watson and Salesforce’ Einstein, for an acceptable price. If these assumptions are correct, Google Jobs may offer recruiters a competitive advantage of several years over Microsoft LinkedIn, Recruit Indeed and Randstad Monster, as well as even more time over Facebook, IBM, Oracle and Salesforce for recruiter applications.

How corporate recruiters, recruitment agencies, and providers of recruiting technology capitalize on this potential competitive advantage depends upon whether you are large (and have access to a large amount of in-house data) or whether you are small (and fleet afoot). Large recruiters with their own proprietary data can do matching beyond what a public-access service like Google Jobs can do.

For example, large employers have performance reviews to identify existing employees in, say, the top quartile of performers for a position based on their experience, skills, behavioral characteristics, personality and related attributes. Google Jobs can calculate a candidate’s commute time but they do not have access to some of the key differentiators that make the candidate a great match for a particular employer.

There also might be different characteristics involved for a staffing agency to match a candidate for a particular open position versus the characteristics that make a candidate a more highly utilized contractor for the agency. A well-run agency knows that their profitability ultimately depends upon how well they can utilize their contractors; they can develop this data internally and Google Jobs doesn’t have access to this data.

If you are a corporate recruiter for a small organization or a recruiter in a placement search or staffing agency that relies on LinkedIn, Indeed or Monster, or if you are using a conventional CRM recruiting and applicant tracking systems that you have used for years, you need to pay attention to the effectiveness of the search and matching tools that you are using. AI and computational linguistics tools are rapidly becoming available through 3 rd party APIs and your recruitment systems vendors should be on top of these changes.

Competition is good and no single competitor will be able to monopolize the data or the search and matching technology. Corporate recruiters and recruitment agencies should be considering how to capitalize on these developments to create a competitive advantage for their own organizations.