Evan Sinar, Ph.D., Manager for Assessment and Selection Analytics, DDI

Online screening tools have recently come under fire as the focus of a June 2012 book by Peter Cappelli, “Why Good People Can’t Get Jobs: The Skills Gap and What Companies Can Do about It.” These tools are blamed for enabling the “purple squirrel” phenomenon where a hiring manager sets the expectation that a single candidate – the purple squirrel – can be found with an ideal combination of skills, education, and experience. As a result, these become must-have qualifications used to filter a candidate pool, often using an automated screening system. Very few candidates emerge from such a search, spawning employer claims of a pervasive skills gap among the available candidate pool. Simultaneously, from the candidate viewpoint, the use of the ideal profile as a benchmark has excluded hundreds or even thousands of possibly suitable individuals from consideration.

Poorly-designed or unstructured applications of online screening tools also have other dangerous consequences, especially when these tools are key workhorses within large-scale recruiting processes. For online screening to fulfill the objectives it was put in place to achieve – improving quality of hire, increasing hiring process efficiency, and democratizing candidates’ access to new job opportunities – users must recognize and avoid several common misuses.

Misuse #1: Hiring Through a Straw, Not a Funnel

Hiring processes work best when designed as a funnel with multiple selection tools ordered in a well-planned sequence in which the full candidate pool enters the mouth of the funnel and candidates move through a series of tools. At each stage, information is captured to screen out those less likely to succeed, with remaining candidates progressed onward until a final hiring decision is made based on the full set of data gathered.

Contrast this model with an online screening approach that is more like a straw than a funnel, where these tools are used to screen out all candidates other than the very low proportion meeting a narrowly-defined profile of skills and experiences, with only this group remaining eligible for hire. Although this straw model can be credited for its focus, this minor advantage is often more than outweighed by its questionable accuracy, excessively constricted definition of suitable candidate backgrounds, and exposure to legal risk, each of which is further outlined below.

Misuse #2: Using the Least Accurate Tools to Eliminate the Most Candidates

Online tools used for initial candidate screening frequently capture three types of information: years of experience, education, and accomplishments indicating skills, knowledge, and training. While this information can aid in prioritizing candidates at the earliest stages of a funnel-based selection process, their relatively low level of predictive accuracy simply does not justify their use as primary screening methods – that is, using these factors alone to exclude a majority of the candidate pool from further consideration.

When ranked against all major selection tools in their ability to predict job effectiveness, measures of training and experience and years of education fall near the bottom, and far below other methods such as cognitive tests, personality tests, and structured interviews. The maximum proportion of candidates to screen out with a single selection tool should roughly match the accuracy with which the tool can gauge job success.

In the case of skill and experience requirements often used for automated candidate screening, their accuracy generally supports screening out a relatively small proportion of candidates, in most cases no more than 20-30%. Be wary about using these factors to screen candidates at substantially higher rates. Rather, supplement and sequence them with other, more predictive, selection tools to strike the right balance between the need to rapidly prioritize a large candidate pool, and the cumulative accuracy for gauging job success of the tools used to reach this goal.

Misuse #3: Codifying a Narrow Route to Job Success

Problems also emerge when online screening criteria are based solely on conventional or “typical” qualifications. Such qualifications may include a minimum number of years performing a similar job, a particular college major, or demonstrated knowledge of a specific industry. When these are used as hard, uncompromising limits to narrow the candidate pool from the outset, a company loses any chance to consider non-traditional backgrounds. If “required” skills and experiences don’t fully account for variations in the experiential opportunities available to those from other cultures and countries, they too can be inadvertently overlooked. Even “cloning” current employees to identify screening factors can be problematic – if a company’s talent status quo is no longer sufficient, modeling new employees after existing ones may only perpetuate current workforce challenges.

The long-term impact of this type of misuse can be wide-ranging given the consequences of narrow employee perspectives for many key talent management issues, including poor global acumen, stifled innovativeness, and process inefficiency stemming from long-unchallenged “standard operating procedures.” Instead, screening factors should accommodate alternative routes to success whenever possible. This could involve broadening the scope of “acceptable” backgrounds to encompass adjacent fields, industries, and skillsets, crediting candidates for their interest in and active learning about emerging areas where extensive formal work experience may not yet be possible, and carefully vetting all criteria to ensure that candidates from different cultures are not mistakenly excluded.

Misuse #4: Generating Legal Risk via Haphazard Screening

The flexibility of online screening systems to enable idiosyncratic filtering of large candidate pools, if not carefully monitored and managed, can be an unrecognized source of legal risk in at least two ways:

First, it can increase the potential for differential treatment of candidates with similar backgrounds. If one such candidate is a member of a legally-protected candidate group and is screened-out while another non-protected group candidate is progressed, this can be considered a form of discrimination.

Second, using an online screening tool to search or prioritize candidates may automatically classify all individuals searched against as formal applicants. This can substantially impact the necessity to document the job-relatedness of each requirement or search term. Although these legal risks can be managed through careful planning, if left ignored they can quickly threaten the defensibility of the entire selection system.

Misuse #5: Neglecting the Untrainables

A final common misuse is factoring a candidate’s past skills, experiences, knowledges, and other “trainables” too prominently into the selection process in general, and within online screening tools in particular. New hire ramp-up time is a critical concern for many employers given lean staffing levels and low budgets available to support extensive training and on-boarding. Hiring an individual who already possesses many of the perceived prerequisites to effective performance is one way to counteract these restrictions.

However, this approach also brings some important caveats. Taken too far it can lead to a situation where “untrainables” – attributes that are difficult or impossible to develop someone in after they have been hired – play, at best, a minor role in progressing candidates to a final interview. These attributes include vital building blocks such as adaptability, integrity, learning adeptness, and innovation, which can be the foundation for an employee expanding his or her role within and beyond the original job description. Hiring processes that fail to early and adequately capture information about these characteristics, using tools designed specifically for this purpose such as cognitive tests and personality inventories, waste the opportunity to understand the “hows” behind a candidate’s background for the sake of knowing only “what” was achieved.

What a candidate brings to the job on day one may indeed help him or her ramp up quickly, but without pairing this knowledge with insights about untrainables, an employer risks bringing on board an employee with a severely restricted ceiling for future growth. Exercise caution to ensure that online screening tools do not overweight trainable characteristics merely because they are more easily searched and screened upon.

Making the perfect the enemy of the good is a critical misstep when online screening tools are used to target the mythical ideal candidate. But it’s far from the only caution to keep in mind, and other consequences of squandering the potential of these tools can be equally damaging or even more so. Online screening is here to stay and can add immense value and efficiency when deployed appropriately. However, if the recent negative attention directed toward online screening systems is causing you to re-evaluate and seek to improve the way you use these tools – and it should – consider this broader set of potential concerns while redesigning your processes and approaches, and when you’re determining the influence online screening tools will have within your broader hiring process.

Evan Sinar, Ph.D., is the Manager for Assessment and Selection Analytics at DDI where he and his team create, demonstrate the ongoing impact of, and continuously evolve DDI’s screening and testing solutions. Follow DDI on Twitter @DDIWorld.