Gajus - Fotolia
LinkedIn's recently announced Glint acquisition is likely to lead to new capabilities and features in Glint's engagement...
platform. The products haven't been announced, but Glint's CEO, Jim Barnett, discussed the general direction in an interview.
This pending Glint acquisition gives LinkedIn another leg in the buildout of its talent management platform, Barnett said. LinkedIn recently announced plans to build an application tracking system, Talent Hub, and just released a new analytics platform, Talent Insights. It had previously acquired a learning platform, Lynda.com.
Glint offers what it calls a real-time platform for tracking employee engagement, in part, by increasing the frequency of engagement surveys, as well the use of pulse surveys. Usually, these surveys are for C-level executives, but Glint's platform shares them with line managers to help them improve their performance.
The processes used in engagement can be expanded to help a firm improve its hiring.
Jim BarnettCEO, Glint
"You could imagine enabling companies to do assessments that would help them understand who is going to be a good fit for a particular role," Barnett said. "We have a platform that would enable that very efficiently."
In its employee engagement product, Glint provides feedback scores, and if it appears a manager can use help in some areas, the system may recommend some actions. This could be as simple as recommending a Harvard Business Review piece to read. With LinkedIn's Glint acquisition, Barnett said he imagines integrating Lynda into Glint.
"The future we see is much more learning embedded in work," he said.
Google for Jobs visibility made easier
In other news, machine learning may be able to help companies navigate Google for Jobs coding requirements. Google for Jobs pulls job ads from around the web -- including corporate job sites -- in response to searches. In doing so, it creates a special page for the job seeker. A Google search, for instance, for a Java developer position will aggregate jobs and offer filters, such as "software engineer" or "senior" to narrow the search.
But Google for Jobs, which arrived in 2017, created complications for corporate job site developers. Google requires a particular tag schema on each job post, which then must be submitted through the Google Index API. This is different from the search engine optimization meta tags that are typically used, said Venkat Janapareddy, CEO of Jobiak, based in Woburn, Mass. His firm developed a tool for addressing this problem.
Janapareddy said most of the job posts with visibility in Google for Jobs were coming from job boards that had adapted to Google's system. But jobs posted on employer job sites were underrepresented, because they weren't coded for it, he said. Some applicant tracking systems support Google for Jobs, but not all.
Every job has to be coded to meet Google's requirements, Janapareddy said. The required tags are job-specific. Jobiak's tool automates this process and uses machine learning to predict all the tags that Google expects, given any job description. The company offers a starter option without charge for firms that want to try it.
The problem with AI in HR
HR's use of AI is hindered by a number of problems, including insufficient data, bias and the difficulty of disentangling individual performance from group performance, according to a new research paper.
The paper, "Artificial Intelligence in Human Resource Management: Challenges and a Path Forward," was written by researchers from the Wharton School at the University of Pennsylvania and the ESSEC Business School in Paris.
The researchers suggested paths for improving HR data analytics, such as "Do not seek perfect measures of performance, which do not exist anyway." Instead, look for reasonable measures, as well as aggregated information from multiple sources over time. They recommended integrating HR data with a company's business and financial data "to analyze effects on business unit performance."
Vendor approaches are compounding the analytics problem for organizations, the researchers said.
"It is very common for an employer to have a system from one vendor to track employee performance scores, from another for applicant tracking software, from a third for compensation and payroll data, and so forth," the paper said.
The paper is available for download at the Social Science Research Network.