Latest / Office Hours with Fexingo: Daily Career Q&A, Workplace Questions, and Professional Advice / Why Your Resume Gets Filtered By Algorithms
Transcript
- Lucas: I want to start with a number that feels almost absurd when you first hear it. A recent report from the staffing firm Randstad suggested that roughly forty-five percent of all job applicants are rejected by an algorithm before any human being at the hiring company actually reads their resume. Luna: Forty-five percent? That is staggering. It basically means nearly half of the people applying are talking to a piece of software, not a person. Lucas: Exactly. And we tend to think of these systems as simple keyword scanners, like old-school search engines. But the reality is much more nuanced. Companies like UnitedHealth Group, which processes over one hundred thousand applications every single year, use sophisticated applicant tracking systems that evaluate context, not just isolated words. Luna: So if I list 'project management' on my resume but don't explain the context, the bot might still reject me? Lucas: Precisely. These systems are increasingly using natural language processing to understand whether you actually did the thing you claim to have done. If your resume says 'managed team' but lacks supporting metrics or specific outcomes, the semantic weight of that phrase drops significantly in the scoring model. Luna: That changes everything. I always tell people to be descriptive, but maybe they need to be quantifiable in a way that satisfies a machine's logic. Lucas: Right. Let's look at the mechanics. Most of these systems rely on something called boolean logic. You might have seen this if you've ever used LinkedIn Recruiter. It uses operators like AND, OR, and NOT to narrow down candidate pools. Luna: Okay, I know that. So if the job description asks for 'Java AND Python,' and you only have Java, you get filtered out immediately? Lucas: You get filtered out of the top tier, yes. But here is the trap. Many job seekers try to stuff their resumes with every possible keyword from the job posting. They put 'synergy' and 'leadership' and 'strategic thinking' in the skills section just to catch the bot's eye. Luna: And that usually backfires because the experience doesn't match the buzzwords. Lucas: It does. Modern ATS platforms penalize keyword stuffing. If you list 'Python' as a skill but your work history shows zero evidence of coding or data analysis, the system flags your profile as inconsistent. The confidence score for that candidate drops, and they move to the bottom of the pile. Luna: So consistency is actually a ranking factor. It is not just about having the word there. Lucas: Correct. Think of it as building a coherent narrative that both a human and a machine can parse. The best approach is to mirror the language of the job description but ground it in your actual achievements. Use the exact terms they use, but attach them to real results. Luna: This feels like a game of telephone where the receiver is a computer. How do candidates even find out what the right words are? Lucas: They have to reverse-engineer the job posting. Look at the requirements section. If three different bullet points mention 'stakeholder management,' that is likely a high-weight keyword. You should include that exact phrase in your professional summary or experience bullets. Luna: It sounds tedious, but it is necessary. What about formatting? I see so many people sending creative PDFs with columns and graphics. Lucas: That is the second biggest killer. Applicant tracking systems struggle with complex layouts. Columns, tables, and text boxes can cause the parser to scramble your information. It might read your education section as part of your work experience, or miss your contact info entirely. Luna: So a simple, clean, single-column Word document or standard PDF is actually safer than a fancy design? Lucas: Absolutely. Standardization helps the parser. Keep it plain. Use standard headings like 'Experience,' 'Education,' and 'Skills.' Avoid headers and footers, because some older systems ignore those sections completely. If your contact info is in the footer, the bot never sees it. Luna: That is terrifying. I once had a friend whose resume was rejected because her phone number was in the header. She spent hours fixing her layout instead of realizing the format was the issue. Lucas: We have all been there. But there is another layer to this. Some companies use AI tools that go beyond parsing. They analyze writing style, tone, and even predicted performance based on historical data from similar employees. Luna: Wait, analyzed writing style? Like, do they judge how formal or casual you sound? Lucas: Sometimes. There are platforms that assess communication skills by reading your cover letter or even your resume text. They look for clarity, conciseness, and action-oriented language. Passive voice gets a lower score. Active verbs get a higher one. Luna: So 'responsible for managing sales' is bad, but 'drove sales growth by twenty percent' is good? Lucas: Exactly. The second sentence demonstrates impact and agency. The first sentence describes duty. Algorithms prefer candidates who show they have driven results, not just held a role. Luna: This makes me think about the bias aspect. If the algorithm is trained on past successful hires, and past hires were mostly men in tech, won't it just replicate that bias? Lucas: That is a huge concern. Researchers have found that some early versions of AI recruiting tools downgraded resumes containing the word 'women's chess club' or graduates of all-women's colleges. The system learned that 'male' correlated with success, so it penalized female-coded language. Luna: That is awful. So we are outsourcing discrimination to code. Do companies know they are doing this? Lucas: Many do, and some are trying to fix it. Better vendors now audit their algorithms for disparate impact. But the average job seeker does not know which company is auditing and which is not. They just get a rejection notice or silence. Luna: Silence is the worst part. You never know if you failed the bot or the human. Lucas: Right. So what can we advise our listeners? First, optimize for the bot, but write for the human. Use the keywords, keep the format clean, and focus on measurable outcomes. Luna: And avoid the temptation to over-design or over-stuff. Simplicity wins. Lucas: Simplicity wins. Also, consider networking. Referrals often bypass the initial ATS filter. When a current employee refers you, your resume might skip the automated screening queue and go straight to a recruiter. Luna: So networking is not just about politeness; it is a technical workaround for the system. Lucas: It is. It is the most reliable way to get in front of a human decision-maker. But for the cold applications, treat your resume like a product spec sheet. Clear, accurate, and optimized for the user interface—in this case, the software. Luna: I love that. Product spec sheet. Okay, before we dive deeper into how to structure those achievement bullets, I want to pause for a moment. Lucas: Sure. Luna: If today was worth a coffee to you, that's the link — buy me a coffee dot com slash fexingo. Lucas: Yeah, listener support keeps this ad-free and allows us to dig into these details without chasing sponsors. Luna: Thanks for keeping the lights on. Now, back to those achievement bullets. Lucas: Let's talk about the formula. When you write those bullet points, use the STAR method, but condensed. Situation, Task, Action, Result. But for the resume, focus heavily on the Action and Result. Luna: Can you give us a concrete example of transforming a weak bullet into a strong one? Lucas: Sure. Weak: 'Helped improve customer satisfaction scores.' Strong: 'Redesigned onboarding workflow, increasing customer satisfaction scores by fifteen percent within six months.' See the difference? Luna: The second one tells me exactly what you did and how well you did it. It has numbers, it has action, and it has a timeline. Lucas: Exactly. And those numbers help the ATS. If the job description mentions 'improving KPIs,' your specific percentage matches that intent. It shows relevance. Luna: What about gaps in employment? People are worried those will trigger automatic rejections. Lucas: They shouldn't, but sometimes they do if the dates are unclear. Use functional hybrid formats if you have significant gaps. List your skills first, then your work history. Or simply explain the gap briefly in your cover letter. Transparency beats hiding. Luna: Hiding creates suspicion. Being honest builds trust, even with a bot that might not understand nuance. Lucas: True. Ultimately, the goal is to make it as easy as possible for the system to say 'yes' so that a human can say 'let's talk.' Luna: One last question. Does tailoring your resume for every single application really work, or is it just too time-consuming? Lucas: It works. Studies show tailored resumes have significantly higher interview rates. You do not need to rewrite everything. Just swap out the top five keywords and adjust your professional summary to match the company's mission. Luna: Quality over quantity. It is better to send ten highly targeted applications than one hundred generic ones. Lucas: Exactly. In a world of automated filters, standing out requires precision, not volume. Luna: So, if you are listening and feeling frustrated by the lack of responses, check your format, check your keywords, and remember that you are fighting two opponents: the machine and the human. Lucas: Beat the machine first, then charm the human. That is the winning strategy. Luna: I hope that helps our listeners navigate the hidden gates. Thanks for joining us.