Is your ATS discriminatory?

With hundreds of applications for every job opening, many companies turn to software to manage the influx and screen for ideal candidates. There are dozens of types of ATS (applicant tracking systems) that can help, but if your HR department doesn't know the ins and outs of how to use them, your software may be working against you.

How does ATS software work?

ATS software is designed to scan resumes and electronic applications to look for keywords and experiences that are most relevant for the open position, matching job seekers to roles and then passing the most relevant resumes to recruiters. In theory, this helps recruiters avoid reading an exhaustive number of resumes to find top candidates. However, as more and more companies rely on ATS software, it may be kicking out ideal candidates in a way that could be construed as discriminatory. 

Ironically, some companies try to customize their ATS to expand diversity in their applicant pool, but it can backfire. Let's dig into how this happens so that you can take steps to hone your own ATS.

How ATS software discriminates and how to fix it

While ATS software is designed to help companies find great candidates, some types have pitfalls and flaws. Applicants can be discriminated against by age, gender, race, and other factors. This not only narrows your pool of job candidates, but it can also be illegal. Here's how it works:

• Ageism and education-based discrimination: By requiring applicants to include their high school or college graduation year, you could be setting up your system for ageism. Requiring that the software automatically decline applicants before a certain year of graduation is ageist. Further, requiring a college graduation year discriminates against applicants who may not have a degree, including those who took some college-level courses but did not finish a degree.
  • Fix it: Don't let the system cut off applicants by requiring a graduation date before a certain year, and don't require college degrees for jobs for which they aren't 100% necessary. Make adding a degree or coursework optional.
    
• Racism. It may come as a surprise, but computer algorithms and software can be biased. Natural language software can prioritize "white-sounding" names, for example. Website features like chatbots can discriminate against candidates for whom English is not a first language. Requiring video screenings or pre-screens can be a huge source of racial discrimination. Even requiring applicants to include their zip code can cause prejudice.
  • Fix it: Eliminate chatbots ( which aren't a personal touch anyway), and don't require one-way videos, which can lead to discrimination.
    
• Gender bias. Don't qualify jobs as "men's" or "women's" jobs either in your system or in an ad. Not only is this discriminatory, but it also excludes candidates who are non-binary. Also, don't use he/she pronouns.
  • Fix it: Have someone from your organization's LGBTQ+ and diversity and inclusion committees review job postings. They can help ensure there aren't gender-skewing words that would turn away people of all genders or make it hard for them to be selected for an interview. If you notice that your job candidates are only of one gender, the software needs fine-tuning.
    
• Ableism. The job application process and job descriptions can be riddled with accessibility barriers. Some examples: Personality tests can discriminate against people who are neurodivergent. Math tests (when math is unrelated to the job) can also discriminate. CAPTCHA tests are intended to be able to tell a human from a robot, but can unintentionally discriminate against candidates with disabilities. 
  • Fix it: Make sure your website and application forms and processes are accessible to everyone, including people using screen readers. Eliminate duplicative steps when possible, such as entering information already provided on the resume. Only list job requirements that are truly essential for the position. Offer alternate ways to apply (such as an email address) if the form is not working for an applicant. Overall, make sure your application process isn't more accessible to one group over another. 

Don't let AI replace human intelligence

If you're using an ATS, make sure your employees are fully trained to use it and can remove aspects that lead to discrimination. Consider having multiple in-house experts review the system to make sure that biases aren't inherent in the setup.

In the end, human intelligence is still superior to computers, so if you're a hiring manager and notice that your candidate pool lacks diversity, your ATS may be doing you more harm than good.