Advertisement
DNA for the Next Generation of Online Screening Systems
Get articles like this
in your inbox
The longest running and most trusted source of information serving talent acquisition professionals.
Advertisement
Related Articles
I view the development of online hiring technology as an evolutionary process. As with all technology systems, online hiring systems are constantly evolving. Systems are created to address specific problems or provide a certain service, and those systems that fail to provide useful solutions or that lack the ability to achieve their stated goals must either introduce some new “DNA” into their gene pool ó or be faced with extinction. On the other hand, those systems with traits that allow them to provide effective solutions will remain alive and continue to leverage their strengths. The purpose of this article is to present some ideas and observations about the DNA that must be introduced into the present genes of online screening and assessment in order to help them evolve and help move the entire online hiring process to its next evolutionary phase. Artificial Intelligence DNA The most critical component of the future survival of online screening and assessment systems is the adoption of artificial intelligence (AI). Innovative AI is the springboard that will allow these systems, and the online hiring processes in which they are embedded, to progress to the next evolutionary level. In the dark ages before the Internet, all screening and assessment measures were delivered in a static manner (e.g., paper and pencil or computer software installed locally on individual PCs). As the infrastructure of the Internet evolved, so did the delivery of screening and assessment measures. This evolution has led to our present state, in which we have the ability to deliver a variety of tests for all kinds of jobs to anyone in the world and score them almost instantly. Make no mistake: this paradigm offers a huge advantage over past delivery infrastructure. However, I would like to make the argument that without the introduction of innovative AI, the present state of evolution for the online screening and assessment paradigm is stuck in the mud. The process of defining of job performance offers a good example of what I am talking about. In order for any selection measure to be both effective and legally defensible, it must measure only traits, skills, etc. that can be shown to be directly related to job performance. The definition of job performance has traditionally been accomplished via a process called job analysis. Job analysis studies can often be complex and extremely time consuming. They are usually done on a localized, case-by-case basis by a small group of qualified (and expensive) professionals using a consultative model. In the past, the need to use experts to conduct job analysis and the complexity of conducting such studies did not impact the usage rate of screening and assessment, because the lack of an efficient delivery mechanism limited the scalability of screening and assessment initiatives. But the Internet has set this model on its ear. Now that old scalability limitations can be thrown out the window, survivability is based on the introduction of systems that combine speed and efficiency. Current demands require systems that can be turned on immediately and can be quickly configured by people in the trenches of the war for talent (i.e. recruiters and hiring managers). These users of online hiring systems don’t have time to wait for job analysis studies, nor do they have the background that has traditionally been required to properly configure screening and assessment systems. The premium on speed and efficiency has left online screening and assessment systems with but one choice: adapt to the need for speed or die. If we fail to provide systems with the ability to quickly and easily define job performance in a manner that provides a legally sound anchor to the online screening process, the ideas (and benefits) of screening and assessment will be passed over for more easily used (but less efficient) methods. For the evolutionary process to favor the use of online screening and assessment, identifying job requirements must become a bulletproof and transparent process that allows unlimited scalability. The only way to accomplish this goal is through the use of new and innovative AI. We must use AI to develop new models that build the intelligence needed to define job performance into the software itself. The creation of such software will require mixing AI DNA into the genes of what we have learned about job performance over the past 50 years. Combining “Matching” and “Measuring” DNA A second critical aspect of the evolution of online screening and assessment is the clear understanding of the difference between “matching” and “measuring” tools and the creation of new systems that combine aspects of both. But what do these terms mean? “Matching” tools are those tools that:
At the present time, matching technology is less than ideal. Most matching is done using resumes and keyword searches, methods that are highly inaccurate and end up wasting the time of both job applicants and organizations. Evolution requires a major change in the current matching paradigm. I define “measuring” tools as those tools designed to measure specific candidate characteristics (like skills, knowledge, abilities, competencies, etc.) that have been demonstrated as being critical for job performance. The results of this measurement can be used to help predict an individual applicant’s ability to perform a specific job. Measurement tools include the tools that I usually refer to as “scientific screening tools” (see my previous article on this subject for more information about these tools). We have over 50 years’ worth of collected data demonstrating that, when used correctly, measurement tools can be highly effective predictors of job performance. While the effectiveness of measurement tools is already well documented, Internet technology and AI provide the mechanisms needed to take their effectiveness to the next level. One of the major pieces of misinformation prevalent today is that matching tools and measurement tools can fulfill the same role and can be used interchangeably. This is not the case. There is presently no technology that can provide a job search process that matches people to jobs using data gathered from measurement tools. Keeping matching and measuring tools separated perpetuates the confusion about their roles and presents a problem for the overall evolution of the online hiring process. The next level of evolution in online screening and assessment requires the blending of DNA from both matching and measuring tools to create new and innovative online hiring systems. Imagine a system, for instance, that can compare stored applicant profiles that include information about experiences, skills, values, traits, and abilities to clearly defined job requirements that have been specified by the hiring organization. Such a system would use a blend of AI and psychometric science to effectively provide matching and measuring at the same time. This, my friends, is evolution at work. I have recently learned about a few innovative systems that are using proprietary AI to take the first evolutionary steps towards marrying the matching and measuring process. Redmatch, for example, offers a product that uses AI to take matching to a whole new level, making keyword searching and resume parsing technology look silly. Guru has been doing innovative things in terms of using AI to blend matching and measuring in a way that has created a whole new paradigm for the job search process. By using AI to begin to blend matching and measuring technology, systems such as these that are providing the first steps down an evolutionary path that will begin creating the DNA needed for building the systems of the future. Customer Service DNA The third type of critical DNA offers the ability to make online job searching a positive experience for the candidate. It seems ironic that in today’s economy, when so many companies are forced to adopt a customer service mentality in order to remain competitive, few if any companies understand the importance of treating applicants as customers. Online job searching is currently one of the most unfriendly processes anyone can be subjected to. The search process entails using a frustrating and inefficient matching tool to locate jobs that are often poorly described and then sending a resume into a black hole. If you are somehow deemed qualified, you may hear back from someone. If you are unqualified, you are left with the feeling that you have been relegated to a giant trash pile in some dank basement. My point is especially clear if you compare online job searching transactions to other online transactions, such as purchases. These transactions work in both directions: the customer is given something of value in return for their time and effort. Those companies that begin to introduce into their genes the DNA needed to make online job searching a two-way transaction that views the job seeker as a customer will gain the competitive advantage needed for survival. Some of the things that I think will be prevalent in the systems of the future include:
Most of these things are not being considered at present because of the large amount of resources needed to develop them. This is another ripe opportunity for AI to automate processes in order to provide the scalability needed to make some of these things a viable option. Once job seekers get a taste of what it is like to be treated properly, they will not stand for doing things the old way anymore. At this point evolution will begin to favor systems that make real value propositions to the job seeker. The Solution: A Whole That Is Greater Than the Sum of Its Parts While each of the three types of DNA I have discussed here are critical to the next phase of evolution for the online job searching process, the real evolutionary leap will come via the combination of the three of them. Evolution requires that screening and assessment content become just one small contribution to a greater entity that leverages a variety of technologies rather than relying on a static delivery system. We already know most of what we need to know to develop assessment content that can accurately predict job performance. In the future, success will be defined by those who are able to take what we know already and mate it with the AI needed to fundamentally alter the genetic makeup of today’s screening systems. This will require developing products using a holistic systems perspective that takes a long-term focus. Of course, these changes will not happen overnight. But an understanding of the goals that must be achieved is required in order for the wheels of progress to be set in motion.