Guide
Resume keywords: finding them, placing them, and not overdoing it
Resume keywords are the specific skills, tools, certifications and job titles a posting names. Take them from the posting's requirements section, use the employer's exact wording, and place each one inside a sentence describing work you actually did. A detached keyword list adds terms without adding evidence.
Published
What counts as a resume keyword
A resume keyword is a specific, checkable term a posting names: a tool, a technology, a certification, a methodology, a job title, a domain. "Kubernetes" is a keyword. "PCI-DSS" is a keyword. "Senior Data Engineer" is a keyword.
What is not a keyword, despite a decade of listicles insisting otherwise: adjectives about you. "Detail-oriented", "results-driven", "team player" and "passionate" are not screened on, cannot be verified, and take up space that could hold a fact.
The test is simple. If a recruiter could plausibly type the term into a search box to filter a pool of applicants, it is a keyword. Nobody has ever searched a candidate database for "hard-working".
How to find the keywords in a job description
The terms are not evenly distributed through a posting. They cluster, and knowing where saves you reading the company boilerplate twice.
- Start with the requirements section. Postings are usually published in a structured format that separates required qualifications from preferred ones. Everything named as required is load-bearing.
- Then read the responsibilities. These describe the work itself and carry the verbs and systems you should be mirroring in your bullets.
- Mark anything that appears twice. A term in both the responsibilities and the requirements is the posting telling you what the role is actually about.
- Note the exact job title. It is what a recruiter searches the pool with, and it is frequently absent from the resumes of people who have done exactly that job under a different internal name.
- Ignore the culture paragraph. The section about being a fast-paced team that loves ping-pong contains no screening terms.
A useful sanity check when you finish: you should have somewhere between eight and fifteen terms. If you have forty, you have included adjectives. If you have three, you have missed the requirements section.
Where to place them so they read naturally
The rule that resolves nearly every placement question: a keyword belongs inside a sentence describing work you did with it. Terms carry weight when they arrive attached to evidence, and lose it when they arrive in a list.
| Placement | Example | Value |
|---|---|---|
| In a skills list only | Skills: Terraform, Ansible, Kubernetes | Matches a search. Proves nothing. Fine as a summary, useless on its own. |
| In a bullet, unquantified | Managed infrastructure using Terraform | Matches, and shows context. Adequate. |
| In a bullet, with outcome | Migrated 40 services to Terraform modules, cutting environment build time from two days to under an hour | Matches, shows context, and gives the human reader a reason to keep reading. This is the version that earns interviews. |
Where the important ones go
- Summary line: the two or three terms that define the role, including the title itself.
- Skills section:the concrete tools, ordered with the posting's named ones first. Keep it short enough to be read.
- First bullet of each role: the keyword most relevant to this posting that you used in that job.
- Nowhere: a detached block of terms at the foot of the page.
Handling synonyms, acronyms and titles
Matching is not always literal, but you should not rely on it being clever. Three cases come up constantly.
Acronyms
Write both forms the first time: "Search Engine Optimisation (SEO)", "Continuous Integration / Continuous Delivery (CI/CD)". Postings are inconsistent about which they use, and spelling out the expansion once costs four words.
Synonyms and near-equivalents
If the posting says "relational databases" and you have PostgreSQL, name both: "PostgreSQL and other relational databases". Do not silently assume the matcher knows they are related — some do, some do not, and a human skimming for their own phrase certainly will not.
Job titles
Never change a title you held. If yours was internal or unusual, add the conventional equivalent alongside it — "Platform Engineer (Site Reliability Engineer)" — which is accurate, searchable, and passes a background check.
Why keyword stuffing loses
Stuffing is adding terms without adding evidence: a wall of technologies at the foot of the page, the same keyword repeated in every bullet, whole phrases from the posting pasted in.
It fails in both passes. Match scores generally weight a term's presence rather than its repetition, so the tenth mention buys nothing the first did not. And the human reading next has seen the pattern many times before — a candidate whose skills list claims nineteen technologies and whose experience section evidences four is answering a question nobody wanted answered.
The economical version of the rule: every keyword should have a bullet behind it. If you cannot write the bullet, the keyword should not be there.
What to do about a keyword you genuinely lack
Some gaps are real. Three honest options, in order of preference:
- Name the adjacent thing you do have. If they want GCP and you have AWS, say AWS. Transferable is a judgement for the reader to make, and they cannot make it if you hid the adjacent experience.
- Address it in the covering note. One sentence acknowledging the gap and what you would bring instead is more persuasive than silence, and far more persuasive than a fabricated claim.
- Apply anyway if the rest matches. Required-qualification lists are frequently wish lists. Missing one item out of eight is not disqualifying; missing six is a signal worth listening to.
What is not an option is listing it. A fabricated skill survives until the technical interview and costs you the relationship as well as the role.
tailorurCV extracts these terms from a posting and reports which your profile already evidences, which is the same method as above with the reading done for you. The full tailoring workflow is in how to tailor a resume to a job description.
Sources
- schema.org — JobPosting — the standard fields a posting is published in — skills, qualifications, responsibilities, occupational category — which is why requirements are so consistently separable
- Google Search Central — JobPosting structured data — documents which of those fields employers are expected to fill in, and therefore where the load-bearing terms reliably appear
- Wikipedia — Applicant tracking system — the candidate-search behaviour that makes exact vocabulary matter to a human recruiter, not only to a matching algorithm
Keep reading
- How to tailor a resume to a job description
A step-by-step method for tailoring a resume to one specific posting: what to change, what to leave alone, and how to tell whether the edit actually helped.
- What an ATS score is, and what it can actually tell you
What an ATS score measures, how match scores are calculated, what counts as a good score, and the limits of any score produced outside the employer's own system.
- How to get past an ATS (and what that phrase gets wrong)
An honest account of what applicant tracking systems do, which tactics survive contact with them, and why the most repeated statistic about ATS rejection is unsourced.
Pull the keywords out automatically
Paste a job description and see the responsibilities, required skills and terms it actually screens on — and which of them your profile already evidences.