The short answer
Resume keywords are the words a recruiter is likely to search for: tools, skills, certifications, job titles and the terms of the field. The posting tells you which ones matter. Use its exact words for things you have actually done, in the places where you did them, once or twice each. Leave out the ones you cannot back with a story.
That is the whole method. A keyword earns its place by being true and by being where a reader expects it. Repetition, long skill lists and hidden text add nothing a search needs, and they cost you with the person who reads the result.
Where keywords come from
There are three good sources, and they work best in this order.
- The posting. Read the requirements, the job title and the description of the work. Mark every tool, skill, certification and domain term. Words that appear more than once are the ones the team cares about most.
- Your actual work. Write down the tools you used, the methods you applied and the names of the things you built or ran. These are the only keywords you can use.
- The standard names in your field. If you used a tool everyone calls by its product name, use that name. "Salesforce" is searchable. "Our CRM" is not.
A source to avoid
Generic keyword lists from the internet are the one source to skip. They tell you what resumes in general contain, not what this employer is looking for or what you have done. A term from a list that matches neither is a line you will be asked about and cannot explain.
Where keywords belong
Place important terms where a reader looks for them: the summary, the skills section, and the bullets that describe the work where you used them. A keyword in a bullet is worth more than the same keyword in a list, because the bullet shows how you used it.
Repeating a keyword once or twice in context beats listing it ten times. Search either finds the word or it does not. The person reading after the search notices the repetition.
Keep the job title your employer actually gave you. If it was unusual, such as an internal name nobody outside the company would recognize, the summary is the place to name the standard role it corresponds to.
Exact match vs. natural language
If the posting says "Project Management Professional (PMP)" and you hold the certification, use that full phrase once. If it says "React" and your resume says "React.js", use the posting's spelling, since both are accurate.
When the posting uses a different word for something you did, use theirs, as long as it describes the same work. If they say "customer retention" and you wrote "keeping clients", the change is accurate and makes you easier to find. If they say "people management" and you mentored one intern, it is not the same work, and the change would be a promotion.
- Spell out an acronym once, then use the short form if space is tight.
- Name the vendor you operated, such as Salesforce, not only the category, such as CRM.
- Match seniority language only when your scope fits: led, managed, owned, supported, contributed.
- Do not add a skill you have only heard of. Interviewers ask about anything prominent on your resume.
A worked example
This example is fictional. The posting is for a Data Analyst at a subscription software company. Reading it twice, Aisha marks these terms: SQL, which appears four times; dashboards, three times; Looker; A/B testing; churn; stakeholders; and, under nice to have, dbt and Python.
Aisha has two years as a reporting analyst. Before touching her resume, she writes down what she actually did.
- I write SQL against our PostgreSQL database every day for revenue and subscription reports.
- I built and maintain the weekly revenue dashboards in Tableau. Finance and product use them.
- I analyzed the results of two pricing A/B tests and presented them to the product and finance leads. I did not design the tests.
- I built a monthly churn report by plan and signup month.
- I have never used Looker or dbt. I did one online Python course and have not used it at work.
Mapping the terms to her notes
Each posting term either matches something in her notes or it does not. The map decides what goes in.
- SQL: daily use. It goes in the skills section and in a bullet that shows what she uses it for.
- Dashboards: yes, in Tableau. The word goes in a bullet, with Tableau named. Looker does not appear anywhere.
- A/B testing: she analyzed tests but did not design them. The bullet says "analyzed", which is what she did.
- Churn: she built the report. The word goes in a bullet.
- Stakeholders: she presented to the product and finance leads, so the term is accurate.
- dbt and Python: not used at work. Both stay off the resume.
Her skills line and bullets, before and after
Before, her skills line read: "Data analysis, Excel, Tableau, Problem solving, Communication, Team player, Fast learner." After: "SQL (PostgreSQL), Tableau, Excel, A/B test analysis, churn and cohort reporting."
Her bullets now use the posting's terms where they describe her real work: "Wrote SQL against the PostgreSQL database daily for revenue and subscription reporting." "Built and maintained weekly revenue dashboards in Tableau used by the finance and product teams." "Analyzed the results of two pricing A/B tests and presented the findings to product and finance stakeholders." "Built a monthly churn report by plan and signup month."
Five of the eight marked terms are now on the page, each next to a piece of work she can talk about. The other three are absent because she does not have them.
What the stuffed version would look like
The tempting alternative is a skills line that reads "SQL, Python, Looker, dbt, Tableau, A/B testing, Snowflake, Airflow". It matches more of the posting. It also lists four tools she has never used, and the first technical question about any of them ends the conversation. A recruiter who needs Looker specifically will not hire her either way. A recruiter who needs a strong SQL analyst will find her through the honest version.
What usually goes wrong
Most keyword mistakes come from treating the resume as something to be matched rather than read.
- A skills cloud of thirty terms. Nobody can tell which ones you use daily and which you saw once.
- Tools you have only heard of. Being on a team that used a tool is not experience with it.
- The posting pasted in white text, or in tiny type. Parsers drop the styling, and the pasted posting appears in your profile for the recruiter to see.
- The same keyword in every bullet. Search counts it once. The reader counts it every time.
- Soft skills as a list. "Team player" and "fast learner" are weak alone, and every resume has them.
- Promoted verbs. "Led" chosen because the posting says "lead", when you supported.
- Ignoring the posting's vocabulary for work you did. If you did the thing, use their name for it.
Soft skills, with evidence
Soft skills are worth showing, not listing. Pair each one with something you did. "Facilitated weekly syncs across design, engineering and sales" shows collaboration better than the word does. "Wrote the onboarding guide new hires still use" says more about communication than "strong communicator".
How Verlane handles it
Verlane treats keywords as questions about your experience, not as words to insert.
- The match score reads the posting into its key terms and checks which ones your resume uses. It counts the usual other ways of saying the same thing, so you are not pushed to copy the posting word for word.
- Matched and missing terms are listed side by side. A missing term is something to check against your experience, not a gap to fill.
- Skill suggestions come with two reasons: why the skill matters for this posting, and where your own material shows it.
- A skill the posting asks for that your material does not show is marked "Confirm you have this". Nothing is added to your resume until you accept it, and a skill you have said you do not have is never suggested.
- The free health check includes a skills evidence check. It points out skills you list that your experience never shows, and suggests showing them in a bullet. It never recommends deleting a skill you have.
- The same health check quotes the buzzwords and clichés it finds, so you can decide which ones to replace with something specific.