When a resume enters a hiring process, the first reader is usually an ATS, the software that compares the text of the document with the terms in the job posting and ranks candidates by how well they match. Those terms are the keywords, and they decide whether the resume ever reaches human eyes.
Choosing them is not a matter of intuition or of generic lists found online. Every posting carries its own keywords, written by the company that is going to filter with them. This guide shows how to extract them from a real posting in four steps, where to place them, and how not to cross the line that separates optimization from forced text.
In this guide
What a keyword is to an ATS
A keyword is any term in the posting that the system looks for on the resume to measure the match between candidate and role. Keywords are not decorative words or adjectives: they are concrete, verifiable nouns.
The categories that weigh most in the score:
- Tools and technologies: Excel, SAP, Google Ads, Figma, Python, AutoCAD.
- Job titles: "digital marketing analyst," "technical sales representative," "backend developer."
- Certifications and credentials: Google Ads certification, CDL license, CPA, PMP.
- Methodologies and processes: Scrum, ISO 9001, consultative selling, omnichannel support.
- Languages with a level: "Spanish (fluent)," "bilingual English/Spanish."
What almost never works as a keyword: "proactive," "results-driven," "team player." Those terms appear on so many resumes that no serious filter uses them to separate candidates.
The 4-step method, with a real posting
The full process takes 10 to 15 minutes per opening and relies on a single piece of material: the text of the posting. Here is a typical example from a Denver-based online retailer, with the keywords already marked in bold:
We are looking for a Digital Marketing Analyst with at least 2 years of experience. Requirements: campaign management in Google Ads and Meta Ads, performance measurement with Google Analytics 4, email marketing (Klaviyo or similar), and intermediate Excel. Preferred: Google Ads certification, e-commerce experience, and familiarity with Shopify.
Step 1: separate requirements from preferred qualifications
Requirements ("campaign management in Google Ads") are mandatory keywords: without them, the score drops below the cutoff. Preferred qualifications ("Google Ads certification") add points but do not disqualify. It helps to mark them in two different colors, because they set different priorities in step 4.
Step 2: underline the concrete nouns
From the text of the posting, pull the tools, titles, certifications, and methodologies exactly as they are written. In the example, ten keywords remain, from "Digital Marketing Analyst" to "Shopify." Everything else (the adjectives, the company culture language) is discarded.
Step 3: prioritize by weight
Not every keyword is worth the same. The ones that appear in the title of the posting, the ones that repeat in the text, and the ones in the requirements list weigh more. In the example, "Google Ads" appears twice (as a requirement and as a preferred certification): it is the central keyword of that opening.
Step 4: cross-check against the resume
With the prioritized list, review the current resume: which keywords are already there, which are written under a different name, and which are missing. The missing ones that are true get added; the ones that are not true never get added, because an invented keyword collapses at the first interview question. This cross-check is exactly what the tool for tailoring the resume to the job automates: it compares both texts and points out the missing matches.
Where to place the keywords
Finding them is half the job. The other half is placing them where the system and the recruiter expect them, because a keyword dropped at the bottom of the document weighs less than the same keyword integrated into the experience.
- Professional title: the line under the name should speak to the title of the posting. For the example opening, "Digital Marketing Analyst" is the obvious choice if it is true to the profile.
- Skills section: the natural home of tools and technologies, with the exact name from the posting ("Google Analytics 4," not "web analytics").
- Experience bullets: the most valuable placement, because it ties the keyword to a fact. "Managed Google Ads campaigns with a monthly budget of $8,000 and reduced cost per lead by 22%" is worth more than the same tool in a list.
- Opening summary: two or three central keywords woven into the profile's introduction.
The remaining formatting conditions (standard headings, a single column, a text-based PDF) are in the guide on how to write an ATS-friendly resume; keywords only work if the system can read the document they sit in.
Natural density or keyword stuffing
There is a temptation to solve the filter by accumulation: repeat every keyword five times, add a block of loose terms at the bottom, or even hide text in white font. All of those tactics share the same problem: modern systems detect them, and whatever the system misses, the recruiter catches on the first read.
A resume that beats the filter with tricks reaches a human who reads an artificial text and rejects it anyway. A repeated keyword adds no extra points in current systems; a keyword backed by an accomplishment does add points in the interview.
The reasonable measure: each important keyword appears once or twice, in the section where it belongs, inside a sentence that would say the same thing even if the ATS did not exist. If reading the resume out loud makes a sentence sound like a list of terms, it is in the wrong place.
Synonyms and variants: cover both forms
Newer systems understand that "Excel" and "spreadsheets" refer to the same thing, but many of the systems still in use do not. Since the candidate cannot know which system is on the other side, the safe strategy is to cover the posting's variant and the most common one, whenever both fit naturally.
| Term in the posting | Variants worth covering |
|---|---|
| Excel | Spreadsheets, Microsoft Excel |
| Customer service | Customer support, client support |
| Google Ads | SEM, paid search |
| Scrum | Agile methodologies, Agile |
| Fluent Spanish | Bilingual (English/Spanish), professional working proficiency |
| B2B sales | Enterprise sales, business-to-business sales |
The practical way to cover both without repeating: one variant in the skills section and the other integrated into an experience bullet. And always with the posting's version as the priority, because the match is calculated against that text.
This work is done per posting, not once. The guide on tailoring the resume to each job explains how to keep that process at 10 minutes per application without rewriting the whole document.
Frequently asked questions
How many keywords should a resume have?
There is no fixed number: it depends on the posting. As a reference, a typical opening contains between 8 and 15 relevant keywords, and a well-tailored resume covers all of the required ones and most of the preferred ones that are true. More important than the count is that each keyword is backed by real experience.
Where do the keywords come from if the posting is very short?
From similar postings for the same role. Searching three or four openings with the same title on job boards and pulling the terms that repeat across all of them gives a reliable list of what the market asks for in that role.
Does it help to put keywords in white font so only the ATS reads them?
No. Current systems extract all text regardless of color, so the trick shows up in the processed profile, and many systems flag it as attempted manipulation. It is a tactic that can only hurt.
Should I write out acronyms or use them as they appear?
Both, when the posting uses one form and the industry commonly uses the other. Writing "Certified Public Accountant (CPA)" or "search engine marketing (SEM)" once covers a recruiter who searches either form. As a rule, the exact spelling used in the posting takes priority.
What if I do not match every keyword in the posting?
Applying can still make sense if the central required keywords are covered. Missing preferred keywords lower the score but do not always disqualify. What makes no sense is adding false terms to compensate: the cost of being found out outweighs any benefit at the filter stage.