How to Tailor Your Resume to Any Job Description With AI (Step by Step)

November 8, 2026 · 17 min read

Quick answer: To tailor your resume to a job description, keep one master resume with everything true about you, then cut a one-page version for each posting: lead each role with the bullets that prove the posting's must-haves, name the skills you have in the posting's own words, and leave out what you can't back up. AI can draft that version in seconds if you lock it to your facts. In our test against 165 live job postings, honest tailoring raised the share of strong matches from 66% to 88%, while a one-line "tailor my resume" prompt made something up in 72% of the resumes it wrote. Check each version with Seera's free resume match checker before you send it.

Image credit: RDNE Stock project / Pexels

How we got these numbers (November 8, 2026): we wrote eight fictional people in eight roles: three analysts (data, business and finance), two engineers (software and data), a product designer, a senior product manager and a customer success manager. Each has a master resume (two dated roles, ten bullets, 13 to 16 skills) and a generic resume, the one-page cut most people send everywhere: the first three bullets of each role and the first eight skills. For each person we drew 25 open English-language postings at random whose title is exactly that role, at most two per employer, from Seera's index of 400,000+ live jobs on employers' own career sites; 165 had a readable skills list. For every posting we wrote the resume six ways and scored each one the way Seera's free resume match checker scores a pasted resume. The AI versions came from a general-purpose AI chat model at its default settings. A separate fact check by the same model decided which of each posting's skills the master resume shows; we compared 40 of its calls from a first run with our own reading and agreed on 38 (both misses were cautious). A skill only counted as made up if it appeared nowhere in the master resume; of 20 flagged rewrites we read by hand, 18 were outright claims and 2 were hedged ("familiar"). That first run, on the same postings the same day, gave strong-match shares within about one point of these for the versions both runs share. No employer is named.

Key Takeaways

• Tailoring works without a single new fact. Reordering and rewording what was already true raised the match level in 26% of checks and lowered it in none.
• The biggest lever was wording. The generic resumes showed a median of 25% of each posting's named skills. Rewritten from the same facts, in the posting's words, they showed 70%.
• A one-line AI prompt is a liability. It claimed a skill the person doesn't have in 64% of rewrites. Fact rules cut that to 6%, but telling the AI to use the posting's exact words pushed it back up to 28%. Do the wording yourself.
• Matching software can't tell tailoring from stuffing. Pasting in skills the person lacked scored about as well as honest tailoring on our own checker. A recruiter and an interview can tell.

Table of Contents

• What tailoring a resume changes, and what it can't
• The test: one person, six resumes, 165 live postings
• How to tailor your resume to a job description, step by step
• Resume keywords: which words to change, and where
• AI resume tailoring: a prompt that keeps to your facts
• Keyword stuffing raised the match too. Here's why it still fails
• Common mistakes
• Frequently Asked Questions
• Useful sources and further reading

What Tailoring a Resume Changes, and What It Can't

A tailored resume is not a different person. It is the same career, cut to one page for one reader: different bullets chosen, a different order, and the posting's vocabulary where yours means the same thing. Nothing gets added that you couldn't explain in an interview.

That matters because a one-page resume can only show part of what you've done. In our test, a fact check found that each person's master resume clearly showed a median of 60% of the skills a posting named, but the checker found only 25% on their generic one-pager. The rest was buried in bullets that didn't make the cut, or written in different words. That gap, between what you have and what the page shows, is the whole job of tailoring.

"One size never fits all. As you apply for jobs, tailor your resume to the position’s requirements. Study the job opportunity announcement and emphasize the parts of your work history that match the qualification requirements listed there."

What tailoring can't change: the gates. If the job needs a work permit you don't have, a licence, or several more years than you have, no wording fixes it, and you should know that before you spend time on a version. Our five-minute eligibility check covers those. And if a must-have is something you have never done, tailoring is the wrong tool; checking your resume against the posting first tells you which gaps are wording and which are real.

The Test: One Person, Six Resumes, 165 Live Postings

For every posting, each fictional person's resume was written six ways:

1. Generic: the one-page cut, unchanged.
2. Tailored by hand: only facts from the master resume. The posting's skills that the person has go first in the skills line, in the posting's words, and each role keeps the three bullets that prove the most of them.
3. Keyword-stuffed: the generic resume plus every skill the posting names that the person does not have, pasted into the skills line.
4. AI, one-line prompt: the master resume and the posting, with "Tailor my resume to this job description."
5. AI, fact-locked prompt: the same, with rules against adding anything. It's printed below.
6. AI, fact-locked plus "exact words": the same rules, plus an instruction to use the posting's exact words for every skill the resume shows, in the bullets and at the start of the skills line. We added it after a first run in which the fact-locked prompt mostly kept the resume's own wording.

Each version was scored like a resume pasted into Seera's checker: strong, good, stretch or weak, from the role, the skills shown, seniority, years, degree and how close the experience reads to the work. Here is the share of postings where each version was a strong match:

Three things stand out.

The more a job title varies, the more tailoring pays. The software engineer gained the most, from 20% strong to 60%. "Software engineer" covers back-end, front-end and infrastructure jobs that ask for very different tools, so a generic page fits few of them; we found the same spread when we checked one resume against 150 postings with that title. The financial analyst, product designer and customer success manager had less room to gain: their generic resumes were already strong for most of their postings.

Honest tailoring never hurt. Across 165 postings, the hand-tailored version never scored a level below the generic one. Cutting bullets that didn't fit the posting cost nothing.

The safe AI prompt was timid, and pushing it made it invent. The fact-locked prompt invented far less than the one-line prompt, but it mostly kept the resume's own words, so it raised the level in 16% of checks against 26% for the hand-tailored version. Adding "use the posting's exact words" lifted that to 21%, and the share of rewrites claiming a skill the person lacks jumped from 6% to 28%. The AI treated the posting's requirement lines as text to copy: one resume gained a skills line reading "Deep Kubernetes Expertise", another a bullet starting "Strong hands-on experience with" a cloud data tool the person had never used.

How to Tailor Your Resume to a Job Description, Step by Step

The first version takes longest. Once your master resume is in shape, most of the work for each new posting is Steps 2 and 6.

Step 1: Keep a master resume

One document, as long as it needs to be, with every role, every bullet worth keeping, every tool you've used in earnest and the numbers behind your results. You never send it. Every tailored version is cut from it, which is what keeps the versions honest: if a line isn't in the master, it doesn't go in a version. Dates matter here too; an undated role can't count toward the years a posting asks for.

Step 2: Pull the job description keywords

Copy the posting from the employer's own careers page and mark three kinds of words in the requirements and the responsibilities:

• Tools and methods named outright: SQL, Figma, Salesforce, A/B testing, financial modeling.
• The work itself, often in the responsibilities: "own the quarterly forecast", "run discovery with customers".
• Level and scope: years, "lead", "mentor", team or budget size.

Split them into must-haves and nice-to-haves. The postings in our test named a median of 8 skills each; most people have more of them than their resume shows, and fewer than they'd like.

Step 3: Mark each must-have as shown, buried or missing

Go down the list with your master resume open. Shown: it's on your one-pager already. Buried: it's in your master but not on the page, or it's there in different words. Missing: you haven't done it. Tailoring works on the buried ones. The missing ones go in your cover letter or stay out.

Step 4: Rewrite the top third

The headline and the first line of the summary get read first, by people and software. Use the posting's job title as your headline if it honestly describes what you do. Rewrite the summary's first sentence to say your role, your years and the two must-haves you're strongest on, in the posting's words.

Step 5: Choose and order the bullets

For each role, pick the three or four bullets from your master that prove the most must-haves, and put the strongest first. Leave the others out of this version; they aren't lost, they're in the master for the next posting. In our test, this selection, plus putting the skills the person had at the front of the skills line in the posting's words, was the whole of the hand-tailored version, and it never scored worse than the generic page.

Step 6: Name your skills in the posting's words, yourself

Go back to your Step 3 list. For each must-have you marked shown or buried, use the posting's term wherever it means the same as yours: in the bullet that proves it and at the start of your skills line. Do this step by hand even if an AI drafted the rest; it's the step where, in our test, the AI started claiming things. The next section shows which changes count.

Step 7: Check the version, then save it

Paste the tailored version and the posting into a resume match checker. If the level moves up and the skills it finds match your list from Step 3, you're done. Save the file with the company and role in its name so you can tell versions apart when the recruiter calls.

Check a tailored resume against its posting, free →

Resume Keywords: Which Words to Change, and Where

"Resume keywords" sounds like a trick. It's closer to translation: the posting has a vocabulary, and your resume should use it for the things you've done.

• ✅ Same thing, their word. You wrote "dashboards in Looker"; the posting says "data visualization". Both are true, so use theirs and keep the tool.
• ✅ Their spelling and their acronym. "Power BI", not "PowerBI"; "user acceptance testing (UAT)" the first time, then "UAT".
• ✅ In the bullet, not only the list. A skills line says you know it. A bullet says where you used it and what came of it.
• ✅ Must-haves first in the skills line, so a skim of the first few words finds them.
• ❌ A near neighbour isn't the same skill. Tableau doesn't make you a Power BI user, and PostgreSQL isn't Oracle. Name what you used, and mention the overlap in your cover letter.
• ❌ No keywords you can't talk about for two minutes. If an interviewer asked about it, could you give an example?

In our test, the hand-tailored versions showed a median of 70% of each posting's named skills, against 25% for the generic pages, with nothing added that the person didn't have. (The checker also credits close variants of a skill, which is why this can exceed the 60% our stricter fact check counted.) That one change in wording did most of the work. For how matching software reads those words, and why a related term can still count, see how AI matches resumes to job descriptions.

AI Resume Tailoring: A Prompt That Keeps to Your Facts

AI resume tailoring is fast and good at the mechanical part: picking bullets, reordering, matching vocabulary. It is also willing to fill gaps with things you never did. Here is what the one-line prompt added across our 165 rewrites:

• A skill the person doesn't have, named in the resume, in 64% of rewrites. Typical additions were whole platforms (Azure, Kubernetes, Power BI, Java) and the posting's softer phrases ("data governance", "functional specifications").
• A results figure that isn't in the master resume in 14%, such as "a 15% reduction in churn" for a churn analysis that never measured one.
• More years of experience than the dates add up to in 2% ("over 7 years" for a five-year career), and "Senior" added to a job title in 4%.
• In 2 rewrites, a line saying the person is authorized to work in the United States. Nothing in the master resume said so.

Every one of those would surface in a background check or the first interview. The fix is to tell the AI what it may and may not do, and to give it the master resume, not your one-pager, so it has real material to choose from. This is the fact-locked prompt we tested:

Tailor my resume to the job posting below. Rules:
1. Use only facts that are in MY RESUME. Do not add any skill, tool, employer, job title, date, degree or number that is not there.
2. Where the posting names a skill my resume clearly shows, use the posting's word for it.
3. Put the experience that matches the posting's must-haves first. Cut what does not help, so it fits on one page.
4. Keep my job titles, employers and dates exactly as written.
5. After the resume, write a line "GAPS:" and list the posting's must-haves my resume does not show. Do not add them to the resume.
Return plain text only.

JOB POSTING:
[paste the full posting]

MY RESUME:
[paste your master resume]

With these rules, the AI named a skill the person lacks in 6% of rewrites and made something up of any kind in 7%, against 72% for the one-line prompt. It raised the match level less often than tailoring by hand (16% of checks against 26%), because it mostly kept your wording. That's the right trade: let the AI choose and order, then do Step 6 yourself. Don't add "use the posting's exact words" to the prompt; in our test that one line took the share of rewrites claiming a missing skill from 6% to 28%.

Lower is not zero, so check the output against your master line by line before you use it:

• Every tool and skill: can you point to where you used it?
• Every number: is it the one in your master, measured the same way? Two one-line rewrites turned "activation rose from 38% to 52%" into "a 14% increase", which is 14 points, not 14%.
• Your name, contact details, titles and dates: unchanged? On one posting, every AI prompt we tried, including one that said "never copy a name, email or phone number from the posting", put the recruiter's name and phone number at the top of the resume.
• The GAPS list: this is your cover-letter material, not your resume's.

Seera doesn't rewrite resumes. Its free resume match checker scores a version against one posting and shows which of the posting's skills it found, and Career Mode ranks your profile against live openings with the missing skills listed per job, which tells you which postings are worth a tailored version at all.

Find your job matches →

Keyword Stuffing Raised the Match Too. Here's Why It Still Fails

We'd rather tell you this than have you find out: pasting the posting's missing skills into your skills line raised the match level in 27% of checks, about as often as honest tailoring (26%), and reached 90% strong. That was on the same scoring our own checker uses. Any tool that reads your resume as text can only see what the text claims; none of them can verify it.

It still fails, for reasons no matching score captures:

• A person reads the shortlist. A skills line full of tools that appear in no bullet reads as a copy of the posting, because it is one.
• The interview tests the claims. "Tell me about a time you used Kubernetes" has no good answer if you haven't.
• The cost lands later. A stuffed resume that gets through costs you an interview slot you'll lose, and possibly your standing with that recruiter for the next role.

Honest tailoring got the same lift without any of that risk. That is the whole argument for doing it the slow way.

Common Mistakes

• ❌ Tailoring from the one-pager. If you or the AI only see the short version, the buried skills stay buried. Start from the master.
• ❌ One version per job title. Postings with the same title asked for different skills often enough that the software engineer's generic page was strong for only 20% of them.
• ❌ Trusting a one-line AI prompt. It wrote in something untrue 72% of the time.
• ❌ Asking the AI to use the posting's exact words. It copied requirements in as experience; claims of missing skills rose to 28%.
• ❌ Changing numbers while rewording. "Rose from 38% to 52%" and "increased 14%" are different claims.
• ❌ Sending without a check. A few seconds in a checker catches a version that lost the skill the posting cares about most.
• ✅ Same facts, new order, their words, checked against the posting before you send it.

Frequently Asked Questions

How do I tailor my resume to a job description with AI?

Give the AI your full master resume, not your one-page version, plus the posting. Tell it to use only facts from your resume, put the matching bullets first and list what is missing separately. Then put the posting's words on the skills you have yourself, check every tool, number, title and date it wrote against your master, and run the result through a resume match checker before you send it.

Can AI tailor my resume without making things up?

Not reliably. In our test, a one-line "tailor my resume" prompt claimed a skill the person did not have in 64% of rewrites and made something up of any kind in 72%. A prompt with explicit fact rules cut that to 7%, but not to zero, and adding "use the posting's exact words" made it worse, so do the wording yourself and read the output against your master resume before you use it.

What are resume keywords, and where should they go?

They are the skills, tools and methods a job posting names. Use the posting's exact words for the ones you have: in the bullet that shows where you used each one, and early in your skills line. Leave out the ones you don't have; a recruiter and an interview will test them even when software can't.

Does tailoring a resume make a difference?

In our test it did. Across 165 live postings, honest tailoring (same facts, reordered, in the posting's words) raised the share of strong matches from 66% to 88% and never lowered a match. The gain was biggest for roles whose postings vary most, like software engineer.

How many tailored versions of my resume should I keep?

One master resume, plus one tailored version for each posting you apply to, named by company and role. If two postings share most of their must-haves, one version can serve both, but check it against each posting first: postings with the same job title often ask for different skills.

Related Reading

• Resume Match Checker: Your Resume vs One Job Description
• AI Job Search That Starts From Your Resume
• How to Check Your Resume Against a Live Job Posting
• How to Match Your Portfolio to a Job Description (With AI)
• How AI Matches Resumes to Job Descriptions, Explained
• Should You Apply? How to Check Your Eligibility for a Job in Five Minutes
• How to Get Better AI Job Recommendations: What We Changed and What Moved
• One-Page PDF Resume With a Portfolio Link, Parser-Tested
• How to Tailor Your UX Portfolio to Each Job Without Rebuilding It
• How AI Resume Screening Tools Score Your CV (and How to Raise Your Score)

Useful sources and further reading

• U.S. Department of Labor: Tips for Writing a Federal Resume — the "Tailor" section quoted above, on matching your work history to the posting's requirements
• O*NET OnLine (U.S. Department of Labor) — standard names for the tasks and skills of each occupation, useful when a posting's term and yours differ
• Seera: Resume Match Checker — free check of one resume against one job description, with the skills it found and missed
• Seera: AI Job Search (Career Mode) — your profile matched against live openings, with written reasons and gaps for each
• Seera tailoring test, run November 8, 2026 — 8 fictional people, 6 versions of each resume, scored against 165 live postings with their exact job title (aggregate counts only; no employer named)
• Seera: Editorial standards — how we use our own data and write about our product