Everyone tells you to copy the JD. Our scorer takes points off for it.

Open any CV advice thread and you'll find the same tip within the first three replies: copy the keywords from the job description into your resume. The ATS is matching words, so give it the words.

It's not wrong. It's just half the story, and the other half is the part that costs people offers.

When we wrote the rubric for Jobr's free CV score, we had to decide what a "good" CV actually means. Not in theory, but as numbers a machine computes the same way every time. Somewhere in that process we added a rule that surprises people: the more distinctive phrases you copy word for word from the JD, the lower your credibility score.

Two readers, not one

Your CV has two readers. The first is software that checks whether the requirements are covered. The second is a person who has read the same JD fifty times today, because they probably wrote it.

When that person sees their own sentence from the posting pasted back at them, it doesn't read as a match. It reads as a candidate who matched the posting instead of describing their work. Lines like "results-driven professional with a proven track record of cross-functional stakeholder alignment" pass the first reader and lose the second.

So our score is split in two. The ATS score measures coverage: how many of the JD's requirements show up in your CV. The credibility score, out of 10, measures whether a human will believe it. It starts at 10 and loses points for:

• every distinctive phrase copied verbatim from the JD (2 points each)
• every self-claim like "proven leader" or "passionate problem-solver" (1 point each)
• every skill you list that never shows up in what you actually did (1 point each)

You can max out the first number and tank the second. Plenty of "optimised" CVs do exactly that.

A skill you never used counts half

The same idea runs through the rest of the rubric. A tool sitting in your skills section, with no bullet where you used it, only gets half credit. A skills list is a claim. A bullet is evidence.

The second version names fewer tools but proves two of them, and gives an interviewer something real to ask about.

We won't add words you can't defend

This is the rule we're proudest of, and it's also the one that makes our rewrites look less flashy than they could.

When Jobr rewrites a CV for a role, every keyword it adds has to pass one test: could you confidently answer "walk me through your experience with this" based only on what was in your original CV? If not, the keyword doesn't go into a bullet. It goes into a list of gaps, so you know it's missing and can decide what to do about it.

You will be interviewed on everything that appears on your CV. A keyword that gets you shortlisted and then sinks you in round one is not a win.

Every rewrite also comes with a short list of the claims you should be ready to back up with specifics in a 30-minute interview. Getting through the filter is the first step. Holding up once you're in the room is the real goal.

Same CV, same score

One last thing we were stubborn about: if you paste the same CV and the same JD twice, you get the same score. Every number is worked out by counting evidence against a fixed rubric, not by a model's impression on the day. A score that moves when nothing changed isn't telling you anything.

What to actually do

Use the JD's vocabulary for skills and tools you've really used. That part of the common advice is right. Then describe the work in your own sentences, put your skills inside bullets instead of a list, and drop any line you couldn't talk about for two minutes.

If you want to see where your CV lands on both numbers, the score is free: paste your CV and a job description into Prep.Jobr. It will also show you which requirements are missing, and the reasons a recruiter would reject you for that specific role.

See how your CV scores against the job

Free ATS score against any job description. 100 free credits when you sign up.