Pull the keywords out of the posting, not out of your head
Copy the full job description into a plain text editor and mark every noun phrase that names a tool, method, certification, domain or deliverable. Those are your candidate keywords.
- Hard skills and tools: Python, Kubernetes, Salesforce, IFRS, Figma.
- Methods and frameworks: A/B testing, Agile, SOC 2, demand forecasting.
- Domain nouns: claims adjudication, supply chain, B2B SaaS renewals.
- Ignore filler: "fast-paced environment", "team player", "rockstar".
Separate required from preferred
Requirements listed under "must have", "required" or in the first three bullets carry the most screening weight. Preferred and "nice to have" items are tie-breakers. Cover every required term you truthfully meet before you spend space on the optional ones.
Match the posting's exact wording — plus one variant
Some parsers do exact string matching. If the posting says "CI/CD" and your resume says "continuous delivery pipelines", a strict matcher scores zero. Write both the way they wrote it and your natural phrasing.
- "Built CI/CD pipelines (continuous integration and delivery) in GitHub Actions."
- Spell out acronyms once: "Search Engine Optimization (SEO)".
- Match singular/plural and punctuation: Node.js vs Node, React.js vs React.
Place keywords where they earn credibility
- Skills section: the flat list of tools, for the matcher.
- Experience bullets: the same terms attached to an outcome, for the human.
- Summary: the three or four terms that define the role you are targeting.
- Never use white text, hidden layers or a keyword block — screening software flags it and recruiters reject on sight.
Aim for coverage, then stop
Above roughly 75% coverage of the meaningful terms in a posting, you are competitive on keywords and further additions start to read as stuffing. Run the resume and the posting through the free checker to see exactly which required terms are still missing.