If you've ever sent out a resume and heard nothing back for two weeks, the first thing worth doing is comparing it against the actual job description you applied for. Not your resume in isolation — your resume against that JD. That's the comparison every recruiter and every parser does in the first 10 seconds, and it's the comparison most candidates skip.
This post shows you the manual method recruiters use (5 steps, takes about 10 minutes per JD), and then the automated path through a tool that does it in 30 seconds. Use whichever fits your workflow. Both produce roughly the same answer if you do the manual one carefully.
The manual method, in 5 steps
This is the same process a careful recruiter does when they screen your resume against a JD. If you can do it on yourself before you submit, you catch the problems they'd reject you for.
Step 1: Pull out the heavy keywords from the JD
Open the job description. Highlight every noun, tool, framework, or qualification that appears more than once, or appears in a "must have" bullet. Those are heavy keywords — they carry more weight in the recruiter's read and in the parser's score.
For a software role, heavy keywords might be "React", "TypeScript", "AWS", "5+ years", "B.E./B.Tech". For a marketing role, "B2B SaaS", "HubSpot", "demand generation", "ABM". Ignore filler like "great communication skills" — those don't move scoring needles.
Step 2: Search each heavy keyword in your resume
For each heavy keyword from step 1, do a literal Ctrl+F search in your resume. Note which ones appear and which ones don't.
The keywords that appear: check that they appear in a meaningful sentence, not just in a skills-section dump. A keyword in a skills list is weak evidence; a keyword in a bullet describing what you did with it is strong evidence.
The keywords that don't appear: these are the gaps. For each gap, ask: do I actually have this experience? If yes, the mirror trick post walks through how to integrate the keyword into an existing bullet without rewriting from scratch.
Step 3: Check role and seniority match
Read the job title in the JD. Then read the title of your most recent role. They should map plausibly.
A "Senior Engineer" applying to a "Staff Engineer" role: fine, the seniority jump is one level. A "Marketing Coordinator" applying to a "VP of Marketing" role: this is going to fail the role-fit check no matter how good the rest of the resume is. If your title doesn't map, the summary section is the only place you can bridge the gap. Use it.
Step 4: Audit the recency window
Read the first three bullets of your most recent role. These get more weight than anything else in the document — recruiters and parsers both treat recent work as the strongest signal of current ability.
Each of those three bullets should map to a responsibility named in the JD, and at least two of them should have a quantified outcome (a number, a percentage, a before/after). If they don't, that's the highest-leverage rewrite you can do today.
Step 5: Check the structure
Final sanity pass: open your resume in plain-text view (most word processors have this option). Read it top to bottom in that view. If sections appear out of order, dates are misaligned, or text from a sidebar or text box ends up scattered across the document, the parser will see the same mess. Fix the structure before anything else; broken structure invalidates everything else you did.
That's the manual method. It takes 10 to 15 minutes per JD if you're being thorough. The output is a list of fixes you can make to your resume, prioritized by signal weight.
The automated method, in 30 seconds
A JD checker tool does the same five steps, just faster. You paste the JD into one box, your resume into another, and the tool returns:
A match score (0 to 100, weighted across the same signals you'd check manually).
The list of missing heavy keywords with suggested integration points in your existing bullets.
A rewritten version of the resume that has the missing keywords worked in without fabrication.
That last one is the practical difference. A good tool doesn't just tell you what's missing; it produces the fixed version. You read it, accept or reject changes, and export.
Resuma does this in about 60 seconds per resume and the free tier covers three full generations, which is enough to see whether the output is useful before you decide to pay for anything. The underlying scoring uses the same 5 signals the manual method audits, so if you've done the manual version once you'll recognize what the score is measuring.
A common mistake: optimizing for the score, not the JD
Tools that score your resume against a JD are useful for the diagnosis. They're less useful as a target.
A score of 95 against one JD usually means you've over-fit your resume to that single posting. The same resume against a similar-but-different JD might score 60. If you're applying to multiple similar roles, the goal isn't to max out the score on any one JD — it's to find a resume version that scores 75+ across the cluster of roles you're targeting.
The practical version of this: don't run the JD checker once and call it done. Run it against three or four similar postings. If your score is high on all of them, the resume is well-tuned for the role type. If it spikes on one and drops on the others, you've over-fit.
What the score actually correlates with
I've watched a lot of resumes go through the Resuma pipeline. The thing the match score correlates with most strongly isn't interview rate — it's whether the resume passed the first screen. That's it.
A score of 85 against the JD says "the recruiter will look at this resume long enough to make a real decision about it". It does not say "you will get the interview". The interview depends on a hundred other things: the strength of the bullets themselves, whether your trajectory makes sense for the role, whether there's a referral in play, what the candidate pool looks like that week.
What the score does is get you past the part of the process that's almost entirely about keyword and signal match. After that, the resume has to be genuinely good. The score can't help you with that part, but the bullet writing and the structure can — start here if your bullets are reading flat.
If you want to skip the manual audit and just see your score against a specific JD in 30 seconds, paste both into Resuma. Three full generations free, no card required. The Inside Resuma's ATS Score post walks through what the five signals are, in case you want to understand the score before you trust it.