How to Write a Data Analyst Resume That Actually Gets Interviews

A report that sat unread is not proof of anything. A query that executed is not proof of anything. Proof is a person, a choice, and a consequence: who took your numbers, what they changed, and what happened after. This guide shows you how to build a resume around that test: what the market looks like right now, the skills the screening software actually matches on, the structure that puts your strongest evidence first, three worked examples, the keywords to mirror, and the mistakes that get these resumes rejected before a human ever sees them.

The market right now: growing on paper, tighter in practice

Start with the honest picture, because the two halves of it point in different directions, and your resume has to be written for both.

The long-term outlook is strong. The Bureau of Labor Statistics projects employment of data scientists, the closest occupation it tracks to a data analyst, to grow 35% between 2025 and 2035, against about 3% for all occupations, with a median annual wage of $120,230 in May 2025 and about 24,800 openings a year. The World Economic Forum's Future of Jobs Report 2025, based on a survey of more than 1,000 employers covering over 14 million workers, names big data specialists the fastest-growing job role, with expected growth of about 113% between 2025 and 2030. One important caveat: the BLS has no separate "data analyst" occupation, so that $120,230 figure covers data scientists too, a broader and higher-paid group. The pay figures closer to the analyst job itself are lower. Huntr's analysis of 65,263 data analyst job posts, current through mid-2026, found the average salary range listed in postings ran about $83,000 to $111,000, and a separate analysis of 2,585 active postings on the InterviewStack job board in May 2026 put the median base salary for US data analyst postings at $87,200.

The short-term picture is tighter. Final Round AI tracked more than 816,000 real interview sessions between June 2025 and June 2026, and data analyst interview activity peaked at 3,431 sessions a month in September 2025, then fell to 1,010 sessions a month by June 2026, a 71% decline, the sharpest contraction of any role on the platform. That is one platform's data, not the whole economy, and it measures interviews rather than postings, so treat it as a signal about the hiring climate rather than a census. But the direction matches what analysts on the ground describe: companies over-hired data professionals between 2021 and 2023, and many are now absorbing the teams they already have. The same tracking found entry-level roles hit hardest, while healthcare was the fastest-growing segment of analyst demand outside tech, and Google, Amazon, 3M, Capgemini, Deloitte, and Microsoft generated the most analyst interview activity. Extern's posting tracker found live data analyst postings rising from July to October 2026, but entry-level postings a shrinking share of hiring in the most AI-exposed occupations, according to Indeed Hiring Lab data from September 2026.

The third force is AI itself. Microsoft Research's 2025 Copilot study ranks data scientists 19th of 785 occupations for AI applicability, and the routine half of the analyst job, pulling numbers, refreshing dashboards, rebuilding the same weekly report, is exactly the half AI now does well. The work that is growing is the judgment half: choosing which metric matters, building a model other people can rely on, spotting where the AI's answer is wrong, and translating a number into a decision. The WEF report expects 39% of workers' core skills to change by 2030, and in analytics that change is already visible in the postings.

For your resume, the message is plain. The market still wants analysts, but it is pickier than it was two years ago, and it is pickier at the junior end than anywhere else. A resume that lists tools and courses reads like every other application in a contracting market. A resume that proves a decision changed gets read.

What the two filters actually check

Your resume passes through two filters before anyone evaluates it. The first is the applicant tracking system, which matches your resume against the keywords in the posting. The second is a recruiter's scan, which TheLadders' 2018 eye-tracking study put at 7.4 seconds. Both filters check the same thing in different ways: named tools, spelled exactly as the posting spells them, and numbers attached to real work.

SQL is the gate. Huntr's 65,263-posting analysis found SQL in 71% of data analyst postings, and required in 61% of them, making it the clearest gate to the job. Python appeared in 46% and split evenly between required and preferred. 365 Data Science's analysis of more than 1,000 postings found the same order at different levels: SQL in 52.9%, Python in 31.2%, R in 24.9%. OneRoadmap read 7,431 analyst postings between late August and mid-October 2026 and found SQL named in 63% of fresher postings and 71% of experienced ones. The InterviewStack analysis of 2,585 postings found the SQL family in 72% and the data visualization and BI family in 74%. The pattern is consistent across every dataset: SQL first, a BI tool second, Python third. A resume that buries SQL, abbreviates it differently, or lists it once in a skills cloud and never proves it in a bullet is failing the first filter.

The second thing the filters check is whether the tools are attached to outcomes. Jobscan's 2026 figures put an applicant tracking system at 97.4% of Fortune 500 companies, and 99.7% of the recruiters in its 384-recruiter survey use keyword filters. But the keyword match is only the entry ticket. The recruiter's six seconds are spent on the first half of page one, looking for a quantified result and a recognizable tool name near the top. Huntr's analysis of roughly 140,000 applications found tailored resumes reached interviews at 4.23%, against 2.07% for untailored ones. Tailoring, in this context, means the posting's exact tool names and the posting's exact metric words, in bullets that prove you used them.

One more filter worth knowing about: verification. Novoresume's 2026 survey found that 86.1% of recruiters had caught resume lies through online checks. A data analyst resume is the easiest kind to check, because a hiring manager can click your GitHub or Tableau Public link and see whether the project exists, whether the code is yours, and whether the numbers in it are real. That link belongs in the header, where it is seen first. For someone with less than three years behind them, it may be the strongest evidence on the page.

The anatomy of a data analyst resume

A data analyst resume has the same sections as any resume, but the order and the content of each section are specific to analytics work. Here is the structure that wins, section by section.

Contact and header. Name, city and state, phone, email, LinkedIn. Then one line that most analyst resumes get wrong by omitting: your portfolio link. GitHub, Tableau Public, or a personal site with two or three projects, in the header where the reader sees it before anything else. No photo, no date of birth, no full street address. For the header, see my guide on should you put your LinkedIn on your resume.

Summary: the field you know, the people you serve, the result you owned. Keep it to three lines. Start with the kind of business you understand, because an analyst who already knows subscription churn, warehouse logistics, or hospital scheduling is useful from week one in a way a generalist is not. Then say who you hand your work to, product managers, finance, operations, and what kind of questions you answer best. Finish with one concrete result. "Data analyst with 4 years in subscription analytics, building churn and pricing models for product and finance teams. Cut voluntary churn 11% by identifying the onboarding step where trial users dropped off, then rebuilding the activation email around it." That is the whole pitch in three lines, and a hiring manager can decide from it in seconds. For the summary-versus-objective choice, see my guide on resume summary vs. objective.

Skills: four clusters, named precisely. Place the technical block just below the summary, where the first scan lands, and organize it so a reader can see the toolkit at a glance. Query languages first, then programming languages, then the visualization platforms, then the warehouses and pipeline tools, then everything else: Excel, version control, statistics, experiment design. Spell each tool exactly as the posting spells it, including the name it carried while you used it. Google's free reporting tool has been renamed twice in four years, and a posting that asks for one name while your resume gives the other can fail a literal keyword match for no good reason. If you used it while it was called Looker Studio, write Looker Studio. A job that ran from 2023 to 2025 says Looker Studio. Small mismatches like that are how keyword screens quietly fail. For the full method, see my guide on how to find resume keywords.

Experience: question, work, consequence. Build each major bullet around three beats, in order. What was the open question. What exactly you did to answer it. What changed because of the answer. "Built funnel dashboard in Looker" is a duty. "Rebuilt the trial-to-paid funnel in Looker after finding that 38% of signups stalled at the payment step; product changed the checkout flow, and trial conversion rose from 11% to 17% in one quarter" is a result. Your current role deserves the most space, three or four bullets. Older roles get fewer, in proportion to how much they say about the job you are chasing now. Every bullet should name a tool, a number, and a decision. If a bullet has a tool but no number, it is a duty. If it has a number but no decision, it is a report. For the verb choice, see my guide on resume action verbs, and for turning duties into proof, see how to quantify resume achievements.

Education and certifications. Put education toward the end for entry and mid-level roles, and follow it with one or two portfolio projects, each with a link and a short description of the data, the question, the approach, and what came out of it. Certificates belong here, with their month and year, and only the ones a hiring manager recognizes: the Google Data Analytics Professional Certificate, the Microsoft Power BI Data Analyst Associate (PL-300), or the AWS Certified Data Analytics credential. The Google certificate is the most common entry-level credential, and it carries real weight with employers: Coursera reports that 75% of US certificate graduates report a positive career outcome, a new job, promotion, or raise, within six months of completion, based on its 2022 program graduate survey, and that more than 150 US employers, including Deloitte, Target, and Verizon, have committed to considering graduates. The certificate costs about $49 a month on Coursera and takes three to six months at a part-time pace. List it, but do not let it stand alone. A certificate with no project behind it is a course. A certificate with a project that changed a number is a credential. For the placement rules, see my guide on how to list certifications on your resume.

Three examples that show the pattern

Examples are worth more than advice here, because the difference between a weak analyst resume and a strong one is visible in two lines. These are fictional composites built from real posting language, so you can see the structure and swap in your own numbers.

Example 1: Entry-level, no direct experience

Priya Sharma | Austin, TX | priya.sharma@email.com | linkedin.com/in/priyasharma | github.com/priyasharma

Data analyst with a B.S. in Statistics and two internship cycles building SQL pipelines and Tableau dashboards. Completed a capstone analyzing 2.1 million retail transactions that identified $340K in revenue leakage. Google Data Analytics Certificate, 2026.

Skills: SQL (PostgreSQL, MySQL), Python (pandas), Tableau, Excel, dbt, Git

Projects

• Retail revenue leakage, 2.1M transactions (Python, SQL, Tableau): joined POS, returns, and inventory data to find that 4.2% of high-value returns were never restocked; built a dashboard flagging the gap, and the store team recovered $340K in a quarter
• Churn prediction, 12,000 subscribers (Python, scikit-learn): built a logistic regression model on usage and billing data; the top-risk decile churned at 41%, against 6% for the rest, and the model now scores new signups weekly

Education: B.S. Statistics, University of Texas at Austin, 2026

A handful of completed projects, each with a write-up a stranger can follow, says more than a long list of half-finished tutorial work. Each one names the dataset, the business question, the method, and the result. The certificate is there, but the projects do the persuading.

Example 2: Mid-level, domain analyst

Marcus Webb | Chicago, IL | marcus.webb@email.com | linkedin.com/in/marcuswebb

Data analyst with 5 years in e-commerce analytics, building pricing and funnel models for merchandising and marketing teams. Owns the weekly revenue forecast; cut forecast error from 9% to 3% by rebuilding the model around promotion calendars.

Skills: SQL (Snowflake, BigQuery), Python (pandas, scikit-learn), Looker, dbt, Airflow, Excel

Data Analyst, Northwind Retail, Chicago, IL | 2022 to present

• Own the weekly revenue forecast for a $40M category; rebuilt the model around promotion calendars and cut forecast error from 9% to 3%
• Built the pricing experiment framework in dbt and Looker; ran 14 A/B tests in a year, and the winning price changes added $1.8M in annual margin
• Automated the Monday reporting pack with Airflow and Python, cutting manual reporting from 6 hours a week to 40 minutes

Junior Data Analyst, Lakefront Logistics, Chicago, IL | 2020 to 2022

• Built the delivery-delay dashboard in Tableau used by 30 dispatchers; on-time delivery rose from 88% to 94% in two quarters

Every bullet names the question, the work, and the decision. The domain is obvious from the first line, and the tools appear in the bullets, not only in the skills block, which is how the keyword screen and the human reader both get what they need.

Example 3: Senior or lead analyst

Elena Ruiz | Seattle, WA | elena.ruiz@email.com | linkedin.com/in/elenaruiz

Senior data analyst with 8 years in healthcare analytics, leading a team of four and owning the metrics layer for a 300-clinic network. Defined the patient-access metric set the executive team now runs the business on; reduced no-show rates 22% by rebuilding the reminder schedule around appointment type.

Skills: SQL (Snowflake), Python, Power BI, dbt, statistics, experimentation, team leadership

Senior Data Analyst, Cascade Health, Seattle, WA | 2023 to present

• Lead a team of four analysts; defined the patient-access metrics that the executive team reviews weekly, replacing 12 conflicting dashboards with one governed set
• Rebuilt the appointment-reminder schedule around appointment type and lead time; no-show rates fell 22% across 300 clinics in six months
• Built the data-quality monitoring in dbt that catches broken pipelines before the morning report runs; cut reporting incidents from 9 a month to 1

Data Analyst, Puget Sound Insurance, Seattle, WA | 2018 to 2023

• Owned the claims-denial analysis; found that 31% of denials came from three coding errors, and the fix recovered $2.4M in the first year

At the senior level, the resume stops being about tools and starts being about judgment: which metric to define, which model to trust, which team to lead. The tools are still named, because the screen still matches on them, but the bullets sell decisions, not dashboards.

The ATS keyword table

Analytics postings are keyword-screened harder than most, because recruiters are usually not analysts and filter on tool names. Here is what to mirror, grouped the way postings group them, with the share of postings that name each one.

Keyword family	What to list	How often postings name it
Query languages	SQL, MySQL, PostgreSQL, Snowflake, BigQuery	SQL in 71% of postings (Huntr, 65,263 posts)
Programming	Python, R, pandas, scikit-learn	Python in 46% (Huntr); R in 24.9% (365 Data Science)
BI and visualization	Tableau, Power BI, Looker, Looker Studio	BI family in 74% of postings (InterviewStack, 2,585 posts)
Warehouses and pipelines	Snowflake, BigQuery, Redshift, dbt, Airflow	Named in most senior and mid-level postings
Spreadsheets and statistics	Excel, A/B testing, regression, forecasting	Excel still named in most postings at every level
Business outcomes	Churn, conversion, forecast accuracy, cost savings, revenue	The words that turn a tool list into a result

Two rules for using this table. First, mirror the posting's exact wording. If the posting says "Power BI" and you write only "BI tools," the parser may miss the match. If it says "Tableau" and you write only "data visualization," the same thing happens. Second, never list a tool you cannot defend in an interview. A hiring manager who uses the same tool will ask one question about it, and the answer will settle the matter. Run the finished file through the free ATS checker to see whether the keywords you meant to include are the ones the parser actually finds, and for the full method see my guide on what "ATS-friendly" really means.

Five mistakes that get these resumes rejected

1. A skills cloud with no proof. Twenty tools in a row, none of them appearing in a single bullet. The parser matches the words, but the recruiter reads the bullets, and a tool that never shows up in your work history reads as a course you finished, not a skill you use. Every tool in the skills block should appear in at least one bullet.

2. Dashboard bullets with no decision. "Built 12 dashboards in Tableau" tells the reader you can use Tableau. It does not tell them whether anyone opened the dashboards, what changed because of them, or whether the work mattered. Add the decision, or cut the bullet. One report that moved a price is worth a dozen reports that sat unread.

3. The Titanic problem. A portfolio of tutorial projects on famous datasets, with no business question attached. Hiring managers have seen the Titanic survival model a thousand times. They have not seen your analysis of a real messy dataset, a real question, and a real result. A handful of finished projects, each with a write-up a stranger can follow, outweighs a long list of class exercises, and a project built on a dataset from your own job, your own city, or a local business is worth more than any Kaggle classic.

4. Claiming statistics you never ran. "Built a machine learning model" when you ran a single regression in a course, or "A/B testing expert" when you never designed an experiment. Analytics is the one field where the interviewer can test the claim in five minutes, with a whiteboard and a question about sample size. Claim only what you have actually run. The parts of the job that are growing are the judgment parts, deciding what to measure and whether to trust the output, and both of those require knowing what you do not know.

5. Hiding the AI question. If you use AI tools to write SQL, draft queries, or build standard reports, say so, and say what you do that the AI cannot: define the metric, check the output, explain the number to a stakeholder. Microsoft's Copilot study ranks data scientists 19th of 785 occupations for AI applicability, which means the routine half of the job is already automatable, and the resume that pretends otherwise reads as out of date. For which AI skills belong on a resume and which do not, see my guide on AI skills on your resume.

No analyst experience yet? Build the proof first

A large share of data analyst hires come from adjacent jobs: customer service, operations, sales, and support, where the person was already the one pulling the numbers. The path between those jobs and an analyst title is shorter than most people think, because the work is the same shape: a question, a dataset, and a decision.

Start with the numbers your old job actually generated. A support agent who tracked ticket volume by category was doing funnel analysis. A store supervisor who built a spreadsheet to predict Saturday staffing was doing forecasting. A sales rep who figured out which leads converted was doing cohort analysis. Write those bullets the way the examples above do, and the hiring manager sees an analyst who has not had the title yet. For the method, see my guide on how to list freelance work on your resume and how to list projects on your resume.

If you have no numbers at all, your first move is not a better resume. It is a project. Pick a real dataset, a public one is fine, ask a question a business would actually pay to answer, and write up the method and the result on GitHub with a short README. Two projects like that, done properly, are worth more than a certificate, because they are the only part of an entry-level resume a hiring manager can verify in one click. The certificate helps, and the Coursera data says three quarters of US graduates report a positive outcome within six months, but the project is what the interviewer asks about.

What does not count is a course with no practice attached. A certificate is fine to list, but it does not replace a single bullet with a number in it. If your resume is a list of courses, stop writing and start building. The numbers are there. They are just not written down yet.

A short note for UK applicants

In the UK, the document is a CV rather than a resume, and the role is usually advertised as "data analyst" just the same. The structure in this guide works the same way, with three differences worth knowing. First, UK hiring managers are more likely to expect the CV to sit on two pages for experienced candidates, and a one-page CV is not a rule the way it is in the US. Second, do not include a photo or a date of birth, and do not include your full address; city and county are enough. Third, salaries are quoted in pounds and the tool mix leans a little more toward Excel and Power BI than Tableau in many UK employers, so mirror the posting's tool names rather than the US defaults. For the full UK picture, see my guide on CV vs resume.

The 30-second audit

Before you send the resume, read it once, out loud, and check five things:

1. The summary names a domain, a stakeholder, and one decision your work changed. If it is all adjectives, rewrite it.
2. Every experience bullet has a tool, a number, and a decision in it. A bullet missing any one of the three is a duty, not a result.
3. Every tool you named is one you used in real work, and every number you quoted is one you can defend in an interview.
4. The keywords from the actual posting appear in your bullets, in the posting's own words, with SQL spelled exactly as the posting spells it.
5. The file is a clean PDF with a single column, your name in the file name, a portfolio link in the header, and no graphics the parser cannot read. For the format question, see my guide on resume PDF vs Word.

FAQ

How long should a data analyst resume be? One page covers most analysts, including those with several years behind them. A second page only earns its keep once you have led a team, owned several specialties, or published work worth listing. Whatever the length, the strongest numbered result goes near the top, because that is where the first scan lands. Leave out old coursework, internships from more than five years back, and any skill-bar graphic.

What is the most important skill to put on a data analyst resume? SQL, by every dataset that counts postings. Huntr found it in 71% of data analyst postings and required in 61%; 365 Data Science found it the most sought-after skill at 52.9%; OneRoadmap found it in 63% of fresher postings and 71% of experienced ones. Put it in the skills block and prove it in a bullet, with the database you actually used.

Do I need a portfolio for a data analyst job? For entry-level roles, yes, and it is often the only proof you have. Two or three projects with a public link, each naming the dataset, the question, the method, and the result, beat a longer list of courses. Put the link in the header, not at the bottom, because it is the one part of the resume a hiring manager can verify in a click.

Is a data analyst resume different from a data scientist resume? Yes, and the difference matters for keyword matching. Analyst postings screen for SQL, a BI tool, and business outcomes; data scientist postings screen for machine learning, modeling, and deployment. The BLS groups the two together in its projections, which is why its 35% growth and $120,230 median figures overstate the analyst job. Write one resume per role, and mirror each posting's exact job title.

Should I list the Google Data Analytics Certificate on my resume? Yes, if you completed it, with the month and year, in the education or certifications section. Coursera reports that 75% of US graduates report a positive career outcome within six months, and more than 150 employers have committed to considering graduates. But list it next to a project, not instead of one. A certificate with no project behind it is a course.

Will AI replace data analysts? Not wholesale, but the role is being reshaped. AI already drafts queries, cleans data, and assembles standard reports, while the BLS still projects 35% growth for data scientists, the closest official occupation, from 2025 to 2035. The routine end of the job, fetching figures and refreshing the same weekly pack, is the part most exposed. The end that is expanding is the thinking end: deciding what to measure, building models other people can trust, catching the AI's mistakes, and explaining what a figure means to someone who has to act on it. Your resume should show the second half, not the first.

The bottom line

A data analyst resume competes against hundreds of identical-looking ones, and the filter that decides it is simple: can the reader see, in the first six seconds, a tool they need, a number that is real, and a decision that followed from it? SQL in the skills block, a BI tool named exactly as the posting names it, and one bullet that says what somebody decided because of your work. That is the whole resume. Everything else is formatting.

The market is still hiring, but it is pickier than it was two years ago, and the entry level is the hardest place in it. The analysts who get through are the ones whose resumes read like evidence, not like a course list. Build the proof, write the decision, and then run your resume through the free ATS checker to see whether the keywords you meant to include are the ones the parser actually finds. If the checker cannot see your tools, neither can the employer's system, and that is a formatting problem worth fixing before you apply anywhere.