Resume Keywords: How to Find and Use Them Honestly
Resume keywords are the skills, tools, certifications and job titles an employer names in a posting. To find them, read the posting, sort its terms into those four groups, and keep only the ones you truly have. Use the employer's wording in your summary, skills list and job bullets. Never add a skill you lack.
This guide covers choosing and placing words. For the whole document, see our guide to making an ATS resume.
Why the employer's exact wording matters
Recruiters can search resumes by word, and the vendors we read say how. Greenhouse's talent filtering page tells recruiters to start by searching for job titles, skills, locations and other keywords "in the full text of resumes and internal notes." It adds: "the keyword from your search must exactly match the keyword in the application" (Greenhouse Support, Talent Filtering, updated March 2, 2026).
Ashby's candidate search offers four match types for resume text. According to Ashby, "Matches" is "a case-insensitive exact match of all the individual words, regardless of order". "Contains" returns records with "the exact phrase, in the exact order." "Similar" returns "records with variations of the word" (Ashby Knowledge Base, Candidate Search). Which option a recruiter picks is their choice.
Two limits apply. We read only Greenhouse and Ashby, so how other systems search is Unknown to us. And a search is not an automatic verdict. Greenhouse says recruiters "might configure a job in Greenhouse to auto-reject or auto-advance applications based on very specific rules". It also says "AI doesn't score or rank applications" (Greenhouse, What really happens after you apply for a job?, January 8, 2025). We found no primary source saying an ATS rejects resumes only because keywords are missing. Our page on what an ATS resume is covers what such systems store and sort.
The practical point is narrow, and it is our reasoning from the two vendors above. If a recruiter types the employer's own term in a search that matches exactly, a resume that uses a different word for the same skill may not appear in that search.
Step 1: Read the posting and mark the terms
Copy the posting into a document. Read it once for the job, then again with a highlighter. Mark every noun phrase that names something you could be asked to prove. Postings repeat what matters most, so note repeated terms. Terms in the title and in a "requirements" list usually carry more weight than a line in the company blurb. That is our reading habit, not a rule from any source.
Step 2: Sort the terms into four groups
Put each marked term in one list:
- Skills: things you do, such as "pipeline reporting" or "customer onboarding".
- Tools: named software or equipment, such as a CRM or a forklift class.
- Certifications and licenses: exact names, such as "CPR certification".
- Job titles: the title in the posting and any titles it mentions for related roles.
Soft-skill phrases ("team player") go in a fifth list you will mostly ignore. Show them through a bullet that describes what you did, not by listing the phrase.
Step 3: Check the vocabulary on O*NET OnLine
O*NET OnLine is a free site that, in its own words, "was created for the general public to provide broad access to the ONET database" of occupations. It was developed for the U.S. Department of Labor by the National Center for ONET Development (O*NET OnLine Help). It lists tasks and tools by occupation. It is not what any one employer wrote, so use it to fill gaps, not to replace the posting.
To look up an occupation:
- Open the O*NET OnLine home page and use the occupation search box, whose label reads "Search by keyword or ONET-SOC code". Type a job title and press the "Search ONET-SOC occupations" button. You can also browse by industry or job family from the same page.
- Open the closest result. We tried Data Scientists, code 15-2051.00. Its summary page has sections for Tasks, Software Skills, Knowledge and Essential Skills, plus Related Occupations.
- Read the Tasks list for how the work is described. The first task listed for Data Scientists is "Analyze, manipulate, or process large sets of data using statistical software."
- Read Software Skills for tool names, which O*NET groups by type with example products.
- Compare with your posting. A term that appears in both is a safe choice if you really did that work. A term only on O*NET is a prompt to ask yourself whether you did it under another name.
The page footer reads "Site updated October 6, 2026."
Step 4: Match only what you really have
Go through each term in your four lists and answer honestly: have I done this, used this, held this?
- Yes, and the posting uses the same words: use its words.
- Yes, but I called it something else: use the employer's term and keep your real meaning. "Client onboarding" in your old job description can become "customer onboarding" if the work was the same.
- Partly: say what you did. Reading reports in a tool is not administering it.
- No: leave it off.
Do not change a past job title to match the posting. A title is a fact an employer can check.
Step 5: Place the terms where a person and a search will both find them
Put a term where it fits the truth, in more than one place when it fits:
- Summary: two or three lines at the top. Use the target job title only if it describes the work you do and want.
- Skills section: a plain list of tools, skills and certifications that you can discuss. List the exact names the posting uses.
- Experience bullets: the strongest place. A bullet shows the term in use: "Built lead-scoring workflows in HubSpot for the sales team."
- Certifications or education: exact certification names and issuers.
A term that appears only in a list tells a reader little. A term in a bullet, with what you did and what changed, lets an interviewer ask about it. Our guide to tailoring a resume to a job description gives a routine for doing this per application.
Step 6: Write the full term and the acronym
A recruiter may type "search engine optimization" or "SEO", and you do not know which. Ashby's "Contains" match returns the exact phrase, and Greenhouse says the keyword must exactly match. Our reasoning: in those searches, one form may not find the other. Write both once: "Search engine optimization (SEO)" in the skills list, then either form in bullets. The same goes for "Certified Public Accountant (CPA)" and "Google Analytics 4 (GA4)". Use the forms the posting uses first.
Worked example
The table below is an invented example, not a real posting. The excerpt reads: "Marketing Operations Analyst. You will manage email campaigns in HubSpot, report on pipeline in Salesforce, analyze traffic in Google Analytics 4 (GA4), and write SQL queries. HubSpot Marketing Software certification preferred."
Imagine a candidate who has worked as a marketing coordinator.
| Posting phrase | Do you have it? | Where it goes | Resume wording |
|---|---|---|---|
| Marketing Operations Analyst | Not as a title. Did campaign operations work | Summary, as a description of focus. Not as a past job title | "Marketing coordinator focused on campaign operations and reporting" |
| HubSpot | Yes, daily | Skills list and one bullet | "Built and maintained lead-scoring workflows in HubSpot" |
| Salesforce | Partly. Reads reports only | One bullet that states the real use | "Pulled pipeline reports from Salesforce for the weekly sales review" |
| Google Analytics 4 (GA4) | Yes | Skills list with acronym, plus a bullet | "Tracked campaign traffic in Google Analytics 4 (GA4) and reported monthly" |
| SQL | No | Leave it off. If you are studying it, say so in one line under Education | (none on the resume) |
| HubSpot Marketing Software certification | Yes | Certifications section, exact name | "HubSpot Marketing Software certification" |
Notice the Salesforce row. The posting wants more than the candidate has, and the wording stays at the true level. The SQL row is empty on the resume on purpose.
Mistakes to avoid
Stuffing. Repeating a term over and over, or pasting the whole posting, makes the resume harder for a person to read. Once a search turns up your resume, a person opens it, and a wall of repeated words costs you trust.
Hidden white text. Words in white font or tiny type are visible to software but not to a reader. We found no vendor page that says how a given system treats hidden text, so we make no claim about detection. The reason not to is simpler, and it is our reasoning: a person opens the file, and a resume that looks padded can end the conversation.
Listing skills you lack. A skill you cannot discuss will surface in the interview. An interviewer can ask for an example of any listed tool, and a test or work sample can follow. Our reasoning: a false keyword may earn a call, and it can cost you the interview or the offer.
Chasing a score. Scanner products publish targets of their own. Jobscan's blog says "We generally recommend a match rate score of 80%" (Jobscan blog, May 13, 2022). That is one company's recommendation for its own tool. The post does not say that employers or ATS software use that score, and we found no ATS vendor stating a match-rate rule. Treat a scanner's score as one prompt to review your wording.
Copying vocabulary from a different job. Keywords from a posting that is not the one you are applying to are noise. Pull them from each posting.
Checklist
- I marked skills, tools, certifications and titles in the posting.
- I checked an O*NET OnLine page for gaps in my vocabulary.
- Every term on my resume is something I can explain in an interview.
- Key terms appear in at least one experience bullet, not only in a list.
- Acronyms are paired with the full term once.
- No white text, no repeated blocks, no false titles.
Sources
All pages were read on 2026-11-12.
- Greenhouse Support, "Talent Filtering": https://support.greenhouse.io/hc/en-us/articles/27104809835291-Talent-Filtering
- Greenhouse, "What really happens after you apply for a job?": https://my.greenhouse.com/blogs/what-really-happens-after-you-apply-for-a-job
- Ashby Knowledge Base, "Candidate Search": https://docs.ashbyhq.com/candidate-search
- ONET OnLine (U.S. Department of Labor, developed by the National Center for ONET Development), home page: https://www.onetonline.org/
- O*NET OnLine Help: https://www.onetonline.org/help/online/
- O*NET OnLine, "15-2051.00 - Data Scientists": https://www.onetonline.org/link/summary/15-2051.00
- Jobscan blog, "What Jobscan match rate should I aim for": https://www.jobscan.co/blog/what-jobscan-match-rate-should-i-aim-for/
Not read: Lever's help pages could not be opened by our tools, and we did not read Workday, iCIMS or SmartRecruiters documentation. How those systems search resumes is Unknown to us.
Disclosure
atsresumeguide.com is published by the makers of AIHawk, a job application tool. The About page gives details.
Last verified: