The keywords an ATS actually looks for are the specific skills, tools, certifications, and job titles repeated in the posting, not generic soft skills like “hard worker” or “team player.” Pull the requirements and responsibilities sections specifically, since that’s where the terms an applicant tracking system weighs most heavily tend to cluster, then match your resume’s wording to theirs as closely as honesty allows. Below is exactly how to pull them out by hand in a few minutes, how to use AI tools for this without producing a resume that reads like a keyword list, and the mistakes that get people screened out even after they’ve done this work.
Contents
- What an ATS Actually Scans For
- How to Pull Keywords Out by Hand
- Using ChatGPT or a Tool Without Overdoing It
- Where to Actually Place Them on Your Resume
- Mistakes That Get People Screened Out Anyway
- Before-You-Submit Checklist
- Frequently Asked Questions
What an ATS Actually Scans For
Applicant tracking systems rank and filter resumes by matching specific terms against the job posting, weighted toward hard skills, tools, certifications, and exact job titles rather than personality traits. A posting that lists “proficiency in Excel, experience with CRM platforms, and a valid driver’s license” is telling you exactly what it’s going to check for; a line like “self-starter with a positive attitude” almost never factors into the actual scan, even though it’s common in postings.
How to Pull Keywords Out by Hand
- Copy the Requirements and Responsibilities sections into a blank document, separate from the rest of the posting.
- Underline every noun that names a specific skill, tool, software, or certification. Ignore adjectives and general phrases for this pass.
- Circle anything repeated more than once, even in slightly different wording. Repetition is usually the strongest signal of what the system is actually weighting.
- Note the exact job title used in the posting, since ATS systems frequently match on title phrasing specifically, not just skills.
- Separate hard requirements from nice-to-haves, usually distinguishable by language like “required” versus “preferred” or “a plus.”
Using ChatGPT or a Tool Without Overdoing It
Pasting a job description into ChatGPT and asking it to list the essential keywords works fine as a starting point, and it’s faster than the manual pass above. The mistake people make is stopping there: an AI tool will often return a long, generic list that includes terms barely mentioned once, and stuffing all of them into a resume produces text that reads like it was written for a machine instead of a person. Use the tool to generate the list, then apply the same “required versus repeated versus mentioned once” filter from the manual method before deciding what actually makes it onto the page.
Where to Actually Place Them on Your Resume
- Job title near the top, matching the posting’s phrasing when your actual title reasonably supports it.
- A dedicated skills section, listing the specific tools and hard skills you underlined, not a restated version of your job history.
- Worked naturally into your experience bullets, attached to something you actually did, rather than listed as a bare word with no context.
- Your summary or headline line, if you use one, weighted toward the two or three terms that showed up most across the posting.
Mistakes That Get People Screened Out Anyway
- Hiding keywords in white text or tiny font. Modern ATS systems generally detect this, and recruiters who do see the raw resume will notice it too; it reads as manipulation rather than qualification.
- Listing a skill you can’t actually speak to in an interview. Getting past the filter only to fail the human conversation that follows wastes everyone’s time, including yours.
- Using a non-standard resume format the system can’t parse. Tables, text boxes, and heavy graphic design can scramble how an ATS reads your content, even when the keywords are technically present.
- Keyword-stuffing to the point the resume no longer reads naturally. A human still reviews resumes that pass the initial filter, and an obviously stuffed document often gets set aside at that stage instead.
- Using the same generic resume for every application. A resume tuned to nothing in particular will underperform one specifically matched to each posting’s actual language, even a fairly close match.
Once your resume is actually landing interviews, our breakdown of when the job market will get better covers what’s driving heavier ATS filtering across the board right now, and our employment verification letter templates are worth having ready once an offer actually comes through.
Before-You-Submit Checklist
- Pulled the specific hard skills, tools, and certifications from the Requirements section
- Noted which terms repeated more than once across the posting
- Matched the job title phrasing where it honestly applies
- Placed keywords in a dedicated skills section and worked them naturally into experience bullets
- Avoided listing anything you couldn’t comfortably explain in an interview
- Checked the resume still reads naturally to a human, not just to a scanner
Frequently Asked Questions
How many keywords from a posting should actually end up on the resume?
There’s no fixed number. Focus on the ones that repeat or are marked as required rather than hitting a target count; a shorter list of genuinely matched terms outperforms a padded one.
Do soft skills in the posting matter at all?
They matter more to a human reviewer later in the process than to the initial ATS scan, which weights hard skills, tools, and titles far more heavily. Don’t ignore them entirely, just don’t expect them to drive your keyword matching.
Do I really need to do this for every single application?
For roles you’re genuinely targeting, yes; a tailored match meaningfully outperforms a generic resume. For a very high-volume, low-effort application strategy, you may reasonably accept a lower match rate on some applications in exchange for applying to more of them.
Does a PDF resume get parsed correctly by an ATS?
Most modern systems handle PDFs fine, though a plain, simply formatted Word document is still the safest choice if you’re unsure which system a specific employer uses.
Do gig or evaluator platforms like Appen use the same kind of ATS filtering?
Generally no. Most evaluator and rater platforms screen through a qualification test or assessment rather than resume-based ATS filtering, so this specific keyword-matching approach matters far less for that kind of application.
