Automated assessmentsMay 14, 2019
Changing how you think about talent acquisition
Organizations have long used a mix of assessments of both skills and behaviors as part of their initial hiring process. These are typically followed by a series of interviews for assessing overall fit (including cultural fit) and are often supplemented by behavioral assessments. However, with the advent of AI, these assessments are rapidly being replaced by more candidate-centric automated options.
Automated assessments are based on assessing candidates in real time and use a scoring/ranking methodology to determine candidate fit for the job and the organization. While much of the burden is on the organization to make sure it uses the right kind of automated assessments for pre-screening candidates, recruiters generally welcome these changes that are helping them prioritize candidates faster, eliminate known biases, recommend best-fit candidates to hiring managers, and, most significantly, focus on creating a more consistent and engaged candidate experience.
Key categories of automated assessments
Chatbots have emerged as a platform for employers to perform basic assessments and act as a first layer of real-time engagement with candidates while requiring less recruiter intervention in the initial stage. These chatbots use natural language processing to facilitate “knockout” questions and evaluate whether candidates meet the minimum job qualifications.
Organizations are also using both skill-based assessments and behavioral/psychometric assessments to determine if candidates “walk the walk.” Both categories are based on machine learning and neural networks, leveraging game-based behavioral analysis and app-based quizzes.
Skill-based assessments provide insights on whether candidates meet minimum requirements or have specific technical skills to excel in their role. In conventional skill assessments, recruiters typically spend 20 to 30 minutes asking about a candidate’s experience, relevant work samples, or offline skill-based tests as part of a pre-screening round. This round is then followed by behavioral and technical interviews with the hiring teams, which has resulted in hiring managers spending time with candidates who may not be technically equipped for the job.
Skill-based assessments with AI algorithms are helping to ensure candidates have the right skills. In particular, the technology, banking, and manufacturing industries are using gamification during the application process. Candidates are directed to take tests for a specific skill set through interactive exercises or mobile-friendly games that evaluate the depth of their knowledge. These platforms can then score candidate responses and rank them for the recruiters. Hence, recruiters are only dealing with high-quality candidates who truly possess the required skills. More recently, passive candidates have also been able to take these assessments via app-based interactive exercises that are scored based on various parameters (e.g., coding skills, problem solving), giving recruiters access to a pool of candidates that was not necessarily looking for a job.
Candidates who have been through an automated skill-based assessment early in the process are far more likely to perform well in the interview process, thus resulting in faster screening times, a reduced number of interviews, and increased credibility of the Talent Acquisition function1.
Also, some organizations have found these assessments highly beneficial to reduce recruiters’ need for technical knowledge of niche skill sets (e.g., a recruiter with limited knowledge of Python can still screen and recommend Python developers to interview). In general, automated assessments have not only demonstrated a track record of improving the traditional metrics like time to hire and cost per hire but have also helped predict the future by providing details on a candidate’s likelihood to jump jobs and predicting success in the role2,3.
Automated behavioral assessments to evaluate a candidate’s cultural fit and cognitive abilities are based on machine learning and neural networks and leverage game-based behavioral analysis and app-based quizzes. Organizations have been using gamification- and simulation-based assessments and virtual reality to evaluate candidates against cognitive performance required for the role. Gamification uses the ability to play games to test key traits and capture different aspects of behavioral assessments through game elements like rules, point scoring, and competition. Simulation-based assessments and virtual reality capture behavioral traits in simulation environments and use predictive models to assess potential candidates on aspects like emotional intelligence and effective leadership.
Use of these techniques has been shown to increase applicant-to-hire ratio and diversity while decreasing the cost and time per hire and reducing the biases that typically come with manual assessments4,5. Customized approaches are also available for organizations to select competencies and attributes that better fit their needs and the role. The behavioral automated assessments are short questionnaires (generally 4-6 questions) aimed at capturing not only candidates’ responses but also their attitudes and emotions through facial recognition. The social and cognitive components of a professional play a key role in job fit and culture fit for an organization.
A few things to consider
A new age of assessments is here today with AI, and many plug-and-play solutions are readily available that integrate with applicant tracking systems and support multiple languages. While a heavy investment is not required up front, organizations are more likely to reap the demonstrated benefits of automated assessments if they have a strong foundational talent acquisition operating model and processes, and a clear understanding of the evaluation criteria for a successful selection.
- Necessary groundwork: The up-front premise of automated assessments is based on organizations being able to clearly define and articulate the role requirements and competencies for hiring talent as well as the values, behaviors, and interactions that drive performance and growth based on the organization’s culture. This exercise typically involves development of an ideal candidate persona and segmenting or enriching the job descriptions for specific positions to reflect the true skills, qualifications, success factors, etc., clearly articulating the organization’s culture and defining expected individual behaviors and interactions.
- Data management: Organizations that do not have mature processes to collect and track data will likely not realize the potential benefits of automated assessments as rapidly as others.
- Desired candidate experience: Organizations should realize that the automated assessments might very well be the first interactive touchpoint the company has with the job applicants. So, it is very importance to make these as personalized and specific as possible.
- Potential for bias: Organizations should proactively address the risk of building unconscious bias in the machine learning component of the tools. Limiting unconscious bias is often portrayed as a value proposition of automated assessment solutions, but machines can learn biases, too, based on business decisions. It is true that machine-learning platforms have the capability to learn about the preferred candidate skills and qualifications, but they can learn from the selection outcomes too. If candidates of a particular age, race, gender, etc., are consistently selected through these assessments, there is a risk of the assessment evaluating candidates on those criteria and assigning them better scores.
The advent of these solutions makes this is an exciting time for the Talent Acquisition function. Being able to augment the recruiting process with chatbots, games, and behavioral simulations makes it possible to accelerate the candidate selection process, bring in the right talent, and truly elevate the role of the function as a strategic adviser to the business
Contributors: Swati Garnaik, Akanksha Arora, and Karnav Shah.
Originally published at Capital H blog