The hard skills for your resume are the specific, checkable abilities the job you want asks for: the software, languages, methods, equipment and licenses named in the posting. Pick them from the posting, not from a list of popular skills, use the posting's own spelling, and keep only the ones you can prove in a bullet and discuss in an interview. Those are the words a recruiter searches for, and they are what screening software checks you against. This post covers how to tell hard skills from soft ones, why the hard ones carry more weight, how to pull them out of any posting, examples by field, and how to prove each one in a single line.
Where the skills go on the page and what order they run in is its own question, covered in our post on the resume skills section. This one is about which skills to choose.
Hard skills versus soft skills
A hard skill is something you can be tested on. Either you can build a pivot table in Excel or you cannot, and an interviewer can find out in five minutes. A soft skill is a way of working, such as communication or leadership, and it shows up in what you did rather than in a test.
| Hard skills | Soft skills | |
|---|---|---|
| Examples | Excel, Python, Salesforce, financial modeling, phlebotomy, AutoCAD, a forklift license | Teamwork, communication, problem solving, leadership |
| How a reader checks it | A named tool, a certificate, a technical question | Evidence in your experience bullets and your interview |
| How it is written in postings | Mostly the same spelling everywhere | Many phrasings for the same idea |
| Searchable by name | Yes | Weakly |
Hard skills come in three useful kinds:
- Tools. Named software, platforms, languages and equipment: QuickBooks, Epic, Jira, React, a CNC machine.
- Techniques. Methods that are not products: variance analysis, A/B testing, root cause analysis, sterile technique, contract redlining.
- Credentials. Licenses and certifications that prove a skill: an RN license, a CPA, a PMP.
The US government's own career database draws the same line. O*NET OnLine, published by the Department of Labor's Employment and Training Administration, offers Software Skills and Soft Skills as separate searches, and each occupation page lists its Software Skills on their own.
Soft skills still matter. In NACE's Job Outlook 2026 Spring Update, 85% of employers said they look for teamwork skills on the resume. The survey had 185 respondents and was sponsored by Jobscan. NACE's own reading is the useful part: "Employers want to see examples." Soft skills for a resume belong inside the bullets that show them, which is covered further down.
Why hard skills carry more weight
Hard skills carry more weight because they are what software and recruiters can check against a posting, and most screening is a check against a posting.
Recruiters search for them by name. Greenhouse documents a Full Text Search that recruiters switch on to search candidate profiles for keywords, with a snippet under each name showing where the terms appear. Its Boolean search help explains AND, OR, NOT, wildcards and quoted phrases, and says nothing about automatic synonyms. A recruiter who types "Power BI" finds the resumes that say Power BI.
Ranking features are built on them. In Greenhouse's Talent Matching, a recruiter adds skills to a calibration and rates each one's importance to the role, and Greenhouse recommends that recruiters focus on 4 to 6 key skills, because the match score is spread across the skills selected. Its Talent Matching FAQ says the feature "compares related terms on each resume to the calibrated skills you set," and that "a longer list of matched terms doesn't always mean a higher match score." Greenhouse also says the feature does not advance or reject anyone on its own. Workday's HiredScore grades candidates A to D for recruiters, and its candidate grades reference ties the grades to qualifications: B means "Candidate meets all basic qualifications" and C means "Candidate meets most but not all basic qualifications." A release note dated 5 August 2026 says the new grading "evaluates each screenable qualification individually against the candidate's parsed resume."
Resume scanners weight them first. Jobscan, which sells a resume match score, says on its homepage that its match rate is based on five priorities in order, starting with hard skills, then education (only when an advanced degree is in the posting), job title, soft skills and other keywords. That is one vendor's model, not any employer's, but it points the same way.
Exact criteria do screen people out. In the 2021 Harvard Business School and Accenture study Hidden Workers: Untapped Talent, 88% of employers believed qualified high-skills candidates were vetted out because they did not match the exact criteria in the job description, rising to 94% for middle-skills workers. That is what employers believe, not a measured rejection rate. It still says the same thing as the documentation above: when the posting names a skill and your resume names it differently or not at all, you are harder to find.
Two lessons follow. Use the posting's own words for skills you have. And do not pad: Greenhouse's own FAQ says a longer list of matched terms does not necessarily raise the score, and every extra item is a question you may be asked.
Find the hard skills a posting asks for
Here is the method on an invented posting. Every name and detail is made up.
Financial Analyst, Larkspur Freight
Build and maintain the monthly forecast model in Excel. Run variance analysis against budget and present results to department heads. Pull data from NetSuite and our SQL warehouse. Build dashboards in Power BI. Requirements: 2+ years in FP&A, advanced Excel, SQL. Nice to have: Power BI, Python, CPA or progress toward one.
Work through it in five passes.
- Underline every noun that is a tool, a method or a credential. Excel, forecast model, variance analysis, NetSuite, SQL, Power BI, FP&A, Python, CPA.
- Split required from preferred. The posting does it for you: Excel and SQL are required, Power BI, Python and the CPA are nice to have.
- Count repeats. Excel appears twice, Power BI twice, SQL twice. Repeated terms are the likeliest to be among the handful a recruiter calibrates or searches for.
- Copy the spelling. The posting says "Power BI", not "PowerBI" or "Microsoft BI". It says "variance analysis", so a bullet that says "budget versus actual review" describes the same work in words nobody will search.
- Sort each one into three buckets.
| Skill | Required? | Your status | What to do |
|---|---|---|---|
| Excel (forecast model) | Required | Built one at your last job | Name it on the skills line and prove it in a bullet |
| SQL | Required | Use it weekly | Same |
| Variance analysis | Implied by duties | Do it monthly, called it something else | Rename the bullet in the posting's words |
| NetSuite | Duties | Used a different ERP | Name the ERP you used; do not claim NetSuite |
| Power BI | Preferred | Built two dashboards | List it and show one dashboard in a bullet |
| Python | Preferred | One course, no work use | Leave it off, or list it as coursework if you could discuss it |
| CPA | Preferred | Not held | Leave it off. Never list a credential you do not hold |
The third bucket matters as much as the first. A required skill you lack is a gap to be honest about, not a word to slip in. If you get the interview, the posting's own list is what the interviewer will work from. For complete resumes built this way for analyst roles, see the data analyst resume samples.
When a posting is thin, or you want to see what a field expects in general, O*NET is a good second source. Its Hot Technologies list counts the software mentioned most often across employer job postings, using Lightcast data. For January to December 2025 it covers 46,885,153 US postings. The top of that list:
| Software | Postings mentioning it, 2025 |
|---|---|
| Microsoft Office software | 3,421,542 |
| Microsoft Excel | 3,211,598 |
| Microsoft Outlook | 1,712,260 |
| Microsoft PowerPoint | 1,618,650 |
| Python | 814,960 |
| Structured query language SQL | 758,250 |
| Salesforce software | 581,755 |
| Amazon Web Services AWS software | 537,989 |
Treat that as a map of the market, not a list to copy. A skill that appears in 3 million postings is not a skill that this posting asked for.
Our post on resume keywords goes deeper on where keywords hide in a posting and when a synonym is safe.
Examples by field
The tools in the second column below appear in the Software Skills lists of the matching O*NET occupation pages, as listed in October 2026. The techniques and credentials are common examples, not a ranking. Use the posting in front of you to decide which ones you need.
| Field (O*NET occupation) | Tools | Techniques | Credentials |
|---|---|---|---|
| Accounting (Accountants and Auditors) | Intuit QuickBooks, Sage 50 Accounting, SAP Concur, Tableau, Alteryx | Account reconciliation, month-end close, variance analysis, audit sampling | CPA, CMA |
| Nursing (Registered Nurses) | eClinicalWorks EHR software, Microsoft Access, Microsoft SharePoint | Patient assessment, medication administration, IV therapy, care plan documentation | State RN license, BLS |
| Software (Software Developers) | GitLab, Atlassian Bitbucket, Spring Boot, Red Hat OpenShift, TensorFlow | Unit testing, code review, API design, CI/CD | Cloud vendor certifications, where a posting asks |
| Marketing (Market Research Analysts and Marketing Specialists) | Salesforce software, Oracle Eloqua, Google Looker Analytics, Tableau, IBM SPSS Statistics | A/B testing, audience segmentation, survey design, SEO | Platform certifications, where a posting asks |
| Project management (Project Management Specialists) | Atlassian JIRA, Asana, Tableau, Procore software, Autodesk AutoCAD Civil 3D | Scheduling, budget tracking, risk registers, stakeholder reporting | PMP |
Three patterns hold across fields. Tools vary by employer even inside one field, which is why the posting beats any table. Techniques are where candidates under-describe themselves, because they did the work without using the name. And credentials are binary: you hold one or you do not, so they are the one place where "related experience" is no substitute. Full samples for two of these fields are in the posts on a software engineer resume and a nursing resume.
Proving a skill in a bullet
A skill on a skills line is a claim. A bullet is the evidence. Every hard skill you list should appear in at least one bullet that says what you did with it, on what, and what came of it.
A simple shape works for almost every field: action, skill, object, result. The result is a real number or a real outcome from your own history. Never invent one.
Invented examples, before and after:
| Weak | Proves the skill |
|---|---|
| Responsible for financial reporting | Built the monthly forecast model in Excel for a 9 region budget and ran variance analysis for department heads |
| Used SQL | Wrote SQL queries against the shipments warehouse to replace a manual weekly export, cutting report prep from a day to an hour |
| Good with patients | Completed patient assessments and medication administration for a 5 bed step-down unit, documenting care plans in the EHR |
| Worked on the website | Ran A/B tests on 3 landing pages in HubSpot and kept the variant that raised sign-ups |
| Managed projects | Kept the schedule and risk register in Jira for a 14 person rollout and reported weekly to the steering group |
Each "after" line does three things. It uses the posting's word for the skill, so a search finds it. It shows the skill at work, so a reader believes it. And it hands the interviewer a question you can answer, which is the point of a resume.
The same shape is how soft skills get onto a resume. "Communication" on a skills line is unverifiable. "Presented the monthly variance results to six department heads" shows it, and it also proves a hard skill at the same time.
Two checks before you send:
- The two minute test. For each hard skill on the page, could you talk about it for two minutes with someone who uses it every day? If not, cut it or label it at its real size, such as "Python (coursework)".
- The bullet test. Search your own resume for each skill on your skills line. If a skill appears only in the list and nowhere in your experience or projects, add the bullet that proves it or take the skill off.
When software picks the skills
Tailoring by hand for every posting is slow, and this is the part of it tools now do for you. ApplyEze, for example, writes a resume for every posting it applies to, and a cover letter when the form asks. It brings the posting's named skills and tools onto your skills line, which means it can put a posting's skill on your page that your own resume did not have. There is no step to remove one before a live run, so use a dry run to see what it writes, and be ready to discuss anything on the page. A certification the posting asks for is not added unless you hold it. Your employers, job titles, dates, schools, degrees and certifications come from your own record, and code rejects a rewrite that adds a figure, employer, school or credential your record does not hold. Our guide to how to tailor a resume walks through the same checks by hand.