Data & AI · Updated 16 September 2026
Data Analyst salary and career roadmap
Answer business questions with SQL, a spreadsheet and a dashboard, fast enough that someone changes a decision because of it.
Salary data checked · outlook from BLS projections released · by Bilal Tahir
- What it is
- A Data Analyst is a person who turns a company's raw database into answers a manager can act on, using SQL, spreadsheets and a dashboard tool such as Power BI or Tableau.
- Salary
- A Data Analyst in the United States earns a median of $68,000 entering the field, $88,000 at mid-career and $115,000 at senior level; the top end is $154,571. (Glassdoor, Indeed and Levels.fyi reconciled against BLS Occupational Employment and Wage Statistics, )
- Outlook
- Employment is projected to grow 12% over the ten years to 2036, with about 7,500 US openings a year. (BLS Occupational Outlook Handbook, 15-2031 Operations Research Analysts, )
- Time to first job
- A career changer starting from zero typically needs 9 to 18 months at 10-15 hours a week to reach a first offer.
- Cost to get in
- The cheapest verified route in (Adjacent-role side door) costs about $0; the most expensive costs about $12,000.
- Roadmap
- Our Data Analyst roadmap is 8 steps and about 630 study hours; the fastest route in is Analytics bootcamp at about 6 months.
- Degree
- No degree is legally required to work as a Data Analyst and a portfolio genuinely substitutes for one at small and mid-size employers, but most large-company postings still list a bachelor's degree as a preference.
- Automation exposure
- Medium. Large language models now write competent SQL and first-draft charts, which compresses the ticket-taking half of Data Analyst work and is one reason Indeed's Hiring Lab counted entry-level postings down 7.5 percent year over year in May 2026. Owning metric definitions, experiments and the stakeholder relationship is what does not automate.
What does a Data Analyst do?
A Data Analyst is a person who turns a company's raw database into answers a manager can act on, using SQL, spreadsheets and a dashboard tool such as Power BI or Tableau. A data analyst turns a messy company database into answers. The same work is also posted as Business Analyst, BI Analyst, Marketing Analyst, Product Analyst, Revenue Operations Analyst.
A data analyst turns a messy company database into answers. The core loop is: a stakeholder asks something vague ("why did signups drop in July?"), you translate it into a query, pull the data with SQL, clean and check it, build a chart or a Power BI/Tableau dashboard, and write two paragraphs that say what happened and what to do. Most of the job is SQL and communication; the modelling is usually descriptive statistics, cohorts, funnels and the occasional A/B test readout. Titles vary a lot: business analyst, BI analyst, marketing analyst, product analyst, revenue operations analyst and healthcare data analyst are all the same skill set pointed at a different domain.
People who thrive are stubborn about data quality, comfortable being the person who says "that number is wrong," and genuinely interested in the business rather than in the tooling. It is the most accessible of the four data roles: no CS degree required, no calculus, and the bar is a working portfolio plus fluent SQL. It is also the one where a domain background you already have (nursing, accounting, logistics, insurance claims, ad ops) is worth more than another certificate.
The honest tradeoff is the entry-level market. Junior analyst is the most crowded rung in data: the Google certificate alone has enrolled 3.8 million people, and Indeed's Hiring Lab found entry-level postings down 7.5% year over year as of May 2026 with hiring tilting hard toward seniority. Pay also caps lower than data science or ML engineering unless you move into analytics engineering, data science, or management. Treat the analyst job as a paid on-ramp, not a destination.
Why Data Analyst pay is high
Analyst pay is not driven by scarcity of the skill - SQL is learnable in a few months - but by the cost of the decisions attached to it. A company running $50M in ad spend, inventory or claims will pay six figures to someone who can find a 2% leak, and the analyst who owns the numbers a leadership team steers by is expensive to replace because the knowledge is company-specific: which table is trustworthy, why revenue in the warehouse differs from the finance close, which metric definition survived the last argument. That is why pay rises steeply with tenure inside one domain (Glassdoor puts the senior average at $132,074 against $63,153 for entry level) while the starting rung stays competitive. It is a role that pays for judgment and context, and only modestly for tools.
What's good
- Lowest barrier to entry of any six-figure-track data role: no degree requirement, no calculus, and a portfolio genuinely substitutes for credentials.
- Skills transfer across every industry, so you can follow the domain you find interesting or the one that pays.
- Mostly predictable hours - 40-45 a week outside month-end close and board-deck season.
- Fast feedback loop: you can see a decision change because of your work within days.
- Clean ladders out: analytics engineer, data scientist, product manager, finance and strategy, or analytics management.
- Remote-friendly relative to most professions, and contract work is readily available.
What's hard
- The most crowded entry-level rung in data; 2025-26 junior hiring is down and referrals matter more than credentials.
- Pay plateaus around $120-150k unless you move into analytics engineering, data science or management.
- A real share of the work is maintenance: fixing broken dashboards, reconciling numbers, answering the same question in a new format.
- Chronic data-quality pain - you will spend more time proving a number is right than producing it.
- Stakeholders often want a chart that confirms a decision already made; saying no is part of the job.
- LLMs are compressing the pure query-writing portion of the role, so ticket-taking analysts are the most exposed.
What a Data Analyst does all day
- 08:45 - Check that last night's dbt or ETL run finished and that the exec dashboard refreshed; chase the data engineer if a source table is stale.
- 09:15 - Stand-up with the marketing or product squad you are embedded in; take two new questions into the queue.
- 09:45 - Write and debug a 60-line SQL query joining orders, sessions and refunds; discover the refunds table double-counts partial refunds and spend 40 minutes confirming it with finance.
- 11:30 - Rebuild a Power BI page because the VP wants the same numbers by region and channel rather than by channel alone.
- 13:00 - Ad-hoc Slack request: 'how many accounts used feature X last quarter?' Turnaround, 25 minutes.
- 14:00 - Read out an A/B test: the lift is +1.4% and not significant; write the memo saying so and defend it in a 20-minute call.
- 15:30 - Document a metric definition in the wiki so the next person does not recompute 'active user' a fourth way.
- 16:30 - Office hours with a stakeholder who wants a dashboard but actually needs a one-off answer; talk them out of the dashboard.
- Monthly - Close-of-month reporting crunch: two or three long days rebuilding and reconciling numbers with finance.
Data Analyst salary in 2026: by level
US, annual, USD. Base plus typical bonus where the source reports it.
A US Data Analyst earns a median of $68,000 entering the field, $88,000 at two to four years and $115,000 at senior level, with the top end at $154,571. These are base-salary figures in US dollars as of September 2026, synthesised from Glassdoor, Indeed and Levels.fyi reconciled against BLS Occupational Employment and Wage Statistics. Pay varies about 15-35% by metro.
There is no separate BLS SOC code for 'data analyst'; BLS folds the work into Data Scientists (15-2051, $120,230 median, May 2025) and Operations Research Analysts (15-2031, $88,940 median, May 2025), both of which skew senior and technical, so market aggregators are the better read at the junior end. Glassdoor (Sept 2026, 22,724 reports) puts the US average at $93,557 with a 25th percentile of $72,276, 75th of $122,260 and 90th of $154,571; entry-level averages $63,153 and senior averages $132,074. Indeed (Sept 6 2026, 6.4k salaries) reports an $86,558 average with a $53,711-$139,495 range, junior at $63,369 and senior at $106,492. Levels.fyi US data analyst total comp runs $68,577 (10th) / $110,000 (median) / $190,000 (90th), with entry level at $80,000 median and senior at $160,000 median - higher because Levels.fyi is skewed to tech employers. Analyst comp is mostly base salary: median equity is $0 and median bonus is $0 on Levels.fyi, with a 5-10% target bonus common in finance, insurance and consulting. Geography matters roughly 15-35%: SF Bay Area, NYC and Seattle sit well above the national median, most of the Midwest and Southeast below it. Fully remote analyst roles increasingly pay a national band rather than a coastal one.
The closest clean BLS proxy is Operations Research Analysts (15-2031): 113,100 jobs in 2025, +12% projected 2025-2035, about 7,500 openings a year. The broader Data Scientists code (15-2051) that absorbs many analyst titles is projected +35% over the same window with 24,800 annual openings from a base of 275,600 jobs. Both are well above the 3% all-occupation average, so demand for the function is not the problem. The problem is the bottom rung: Indeed's Hiring Lab reported entry-level postings down 7.5% year over year as of May 2026 and down 6.3% versus January 2025, with the labour market tilting toward seniority (in software development, senior roles were 69.3% of Q1 2026 postings against 4.5% entry level). Automation risk is genuinely medium rather than low: LLMs now write competent SQL and first-draft charts, which compresses the 'pull this number' part of the job. What does not automate is knowing which table is right, which stakeholder is asking the wrong question, and what to do about the answer. Analysts who stay at ticket-taking SQL are exposed; analysts who own metric definitions, experiments and the business relationship are not.
“Employment of operations research analysts is projected to grow 12 percent from 2025 to 2035, much faster than the average for all occupations.”
Data Analyst salary by city
National bands scaled by metro wage differentials from the BLS May 2025 OEWS release.
Data Analyst pay is highest in San Jose / Silicon Valley (mid-career median about $124,960, ×1.42 the national figure) and lowest among large metros in Salt Lake City (about $83,600). The multiplier moves the offer, not what you keep after rent and state tax.
| Metro | Entry | Mid | Senior | vs national |
|---|---|---|---|---|
| San Jose / Silicon Valley | $96,560 | $124,960 | $163,300 | ×1.42 |
| San Francisco Bay Area | $91,800 | $118,800 | $155,250 | ×1.35 |
| New York City | $87,040 | $112,640 | $147,200 | ×1.28 |
| Seattle | $81,600 | $105,600 | $138,000 | ×1.2 |
| Boston | $78,200 | $101,200 | $132,250 | ×1.15 |
| Washington DC metro | $76,160 | $98,560 | $128,800 | ×1.12 |
| Los Angeles | $73,440 | $95,040 | $124,200 | ×1.08 |
| Chicago | $71,400 | $92,400 | $120,750 | ×1.05 |
| Austin | $71,400 | $92,400 | $120,750 | ×1.05 |
| San Diego | $70,720 | $91,520 | $119,600 | ×1.04 |
| Denver | $69,360 | $89,760 | $117,300 | ×1.02 |
| Philadelphia | $69,360 | $89,760 | $117,300 | ×1.02 |
| Dallas | $68,000 | $88,000 | $115,000 | ×1.0 |
| Minneapolis | $68,000 | $88,000 | $115,000 | ×1.0 |
| Raleigh-Durham | $68,000 | $88,000 | $115,000 | ×1.0 |
| Houston | $67,320 | $87,120 | $113,850 | ×0.99 |
| Atlanta | $66,640 | $86,240 | $112,700 | ×0.98 |
| Phoenix | $64,600 | $83,600 | $109,250 | ×0.95 |
| Miami | $64,600 | $83,600 | $109,250 | ×0.95 |
| Salt Lake City | $64,600 | $83,600 | $109,250 | ×0.95 |
A multiplier raises the number on the offer letter, not what you keep. San Francisco pays about 35% more than the national median for these roles, but median Bay Area rent and California state income tax eat most of that for anyone below the senior rung. Texas, Florida, Washington, Tennessee and Nevada levy no state income tax, which is worth roughly 4-10% of take-home versus California or New York City, where city tax stacks on top of state tax. Run the comparison on after-tax income minus housing before you move. Fully remote roles are the edge case worth chasing: a national pay band spent in a 0.88 cost market beats a 1.35 salary spent in a 1.6 cost market for most people.
How the multipliers are derived
Anchor: the BLS May 2025 OEWS national mean wage across all occupations is $33.54/hr ($69,770/yr). Metro all-occupation means from the same release: San Jose-Sunnyvale-Santa Clara $57.32 (1.71x national), San Francisco-Oakland-Fremont $48.19 (1.44x), Washington-Arlington-Alexandria $44.20 (1.32x), Seattle-Tacoma-Bellevue $44.13 (1.32x), Boston-Cambridge-Newton $43.09 (1.28x), New York-Newark-Jersey City $41.50 (1.24x), Denver-Aurora-Centennial $39.28 (1.17x), Atlanta-Sandy Springs-Roswell $34.57 (1.03x), Chicago-Naperville-Elgin $34.42 (1.03x, May 2024), Dallas-Fort Worth-Arlington $33.96 (1.01x), Phoenix-Mesa-Chandler $33.48 (1.00x). We damp the top end of those raw ratios. All-occupation means exaggerate the gap for the careers on this site, because national pay bands, remote hiring and company-wide equity grids compress geographic spread for high-skill professional roles more than they do for service and hourly work. Levels.fyi shows the same damping: its Bay Area software engineer average total compensation of about $291k sits roughly 1.3x-1.4x the US median, not 1.7x. Non-US multipliers convert local market rates to USD and are directional, not survey-grade.
How to become a Data Analyst
Every route we could verify, with honest time, cost and difficulty.
There are 5 routes we could verify into Data Analyst work. The fastest is Analytics bootcamp at about 6 months; the cheapest is Adjacent-role side door at about $0. Internal transfer from a non-analyst job at your current employer produces more career-changer hires than cold applications do, because it converts domain knowledge you already have into the thing employers are actually short of; the certificate-plus-portfolio route works too, but it is the slowest of the five, at about 14 months and 150 to 300 applications.
Internal transfer from your current job
By far the highest-conversion path for career changers. Take the reporting work nobody wants at your current employer - the weekly ops report, the claims reconciliation, the marketing spend sheet - rebuild it in SQL and Power BI, and ask for the analyst title once you are already doing the job. You keep the domain knowledge that makes you employable and skip the resume screen entirely.
Certificate plus portfolio, applying cold
Google Data Analytics or IBM Data Science certificate for structure, then three portfolio projects on real public data with written business conclusions, then 150-300 targeted applications into one or two industries. Works, but slowly: expect 6-12 months of searching in the 2026 market, and expect referrals to matter more than the certificate.
Adjacent-role side door
Take a job that touches data but hires more easily - operations coordinator, customer support analyst, revenue operations, clinical research coordinator, underwriting assistant, ad ops - and move into analytics from inside in 12-18 months. Lower starting pay, dramatically higher hit rate than cold applications.
Contract / staffing agency
Agencies (Robert Half, TEKsystems, Insight Global and similar) place junior BI and reporting analysts on 6-12 month contracts with a lower bar than direct hire. Pay is hourly with thin benefits, but a completed contract is real experience on a resume and contract-to-hire conversion is common.
Analytics bootcamp
A 3-6 month bootcamp with a real capstone and interview coaching. Only worth the $8,000-$16,000 if it forces you through projects and mock interviews you would not do alone, and only with a program that publishes audited outcomes. The credential itself carries little weight; the deadline pressure and the alumni network are what you are buying.
Data Analyst roadmap: 8 steps, 630 hours
Becoming a Data Analyst from zero takes about 630 study hours across 8 steps, roughly 9 to 18 months at 10-15 hours a week plus a job search. Step one is Get fluent in SQL before anything else. SQL comes first because it is the round you get filtered on, and over 80 percent of Data Analyst postings name it. Portfolio projects come last because they are what differentiates you once you are past the screen, and they are only convincing if you can already write the queries behind them.
0 of 8 steps done · saved in this browser ·
-
1
SELECT, WHERE, all four JOIN types, GROUP BY with HAVING, subqueries, CTEs, window functions (ROW_NUMBER, RANK, LAG, SUM OVER), date truncation and date math. Get to the point where you can write a 40-line query with three CTEs without looking anything up.
Why now: Over 80% of data analyst postings name SQL, and almost every hiring process starts with a live or timed SQL screen. Nothing else you learn matters if you fail that round. This is the single highest-leverage block of study in the whole roadmap.
- course Associate Data Analyst in SQLDataCamp · 39 h · Included in DataCamp Premium, about $28/mo billed annually
- practice SQL 50 study plan and the Database problem setLeetCode · 30 h · Free; LeetCode Premium optional at about $35/mo
- practice SQL interview questions by companyDataLemur · 20 h · Free tier; paid tier for the full set
-
2
Pivot tables, XLOOKUP/INDEX-MATCH, absolute references, data validation, scenario tables. Then the statistics an analyst actually uses: mean vs median vs mode and when each lies, variance and standard deviation, distributions, correlation vs causation, sampling error, confidence intervals, and what a p-value does and does not tell you.
Why now: Half of the analyst job still happens in Excel or Sheets - over 60% of postings name them - and finance, ops and healthcare teams live there. The statistics keeps you from confidently reporting noise, which is the fastest way to lose credibility in your first year.
- course Google Data Analytics Professional Certificate (courses 1-5)Coursera · 90 h · $49/month or included in Coursera Plus
- book Practical Statistics for Data Scientists, 2nd EditionO'Reilly (Bruce, Bruce & Gedeck) · 25 h · About $60 print, or included in an O'Reilly subscription
-
3
Power BI or Tableau - not both. For Power BI: Power Query for transformation, a proper star-schema data model, DAX measures (CALCULATE, filter context, time intelligence), row-level security, and publishing to a workspace. For Tableau: LODs, calculated fields, parameters, dashboard actions.
Why now: Dashboards are the deliverable most analyst jobs are actually hired to produce, and 'can build a model and write DAX' separates you from the large pool who can only drag fields onto a canvas. Choose Power BI if you are targeting corporate, healthcare, finance or anything Microsoft-shop; Tableau if you are targeting tech and consumer brands.
- cert Microsoft Certified: Power BI Data Analyst Associate (Exam PL-300)Microsoft · 60 h · $165 exam fee in the US
- course PL-300 free study path and practice assessmentMicrosoft Learn · 40 h · Free
-
4
pandas (read, merge, groupby, pivot, reshape), matplotlib or seaborn for quick charts, requests or an API client for pulling data, and enough scripting to automate a recurring report. Jupyter notebooks as your working environment. You do not need object-oriented Python or algorithms.
Why now: Python is named in roughly half of analyst postings and is the difference between a reporting analyst and one who can be trusted with a real investigation. It is also the on-ramp to the data scientist path if you want it later.
- course IBM Data Science Professional Certificate (Python, pandas and databases courses)Coursera · 120 h · Free to audit; certificate included in Coursera Plus ($59/mo or $399/yr)
- book Python for Data Analysis, 3rd EditionO'Reilly (Wes McKinney) · 35 h · Free to read online; about $65 print
-
5
Not 'I cleaned the Titanic dataset'. Three end-to-end analyses on real public data, each with: the business question, the SQL, the chart, and a written recommendation with a dollar or percentage figure attached. Good sources are public retail and e-commerce transaction sets, CMS healthcare claims, city 311 and transit data, and the NYC taxi dataset. Publish each as a GitHub repo with a README written for a manager, plus a live dashboard link.
Why now: This is what actually converts applications into interviews for career changers with no degree. Recruiters read a portfolio as proof of execution. Titles in your projects should match the jobs you want - if you want e-commerce analyst roles, do e-commerce analyses.
- practice Kaggle Datasets and NotebooksKaggle · 40 h · Free
- book Storytelling with Data: A Data Visualization Guide for Business ProfessionalsWiley (Cole Nussbaumer Knaflic) · 12 h · About $35
- practice Data ProjectsDataCamp · 30 h · Included in DataCamp Premium
-
6
Pick one: healthcare (claims, ICD/CPT codes, HEDIS measures), finance (P&L structure, AR/AP, unit economics), marketing (CAC, LTV, attribution, MER), e-commerce (conversion funnel, basket size, returns), supply chain (fill rate, lead time, safety stock), or insurance (loss ratio, reserving). Learn the 30 metrics that matter and how they are defined.
Why now: Domain knowledge is the cheapest differentiator available to a career changer, and it is the thing an equally-certified competitor does not have. Healthcare and financial risk analytics were the strongest analyst-hiring segments going into 2026. If you already have a domain from a previous career, use it - that background plus SQL beats a CS degree with no context.
- practice Read 30 job postings in your chosen domain and list every metric and tool namedLinkedIn Jobs / Indeed · 8 h · Free
- course Google Advanced Data Analytics Professional CertificateCoursera · 153 h · $49/month or included in Coursera Plus
-
7
Do 60-80 timed SQL problems until window functions are automatic. Then practise the case format out loud: 'signups dropped 12% last month, how would you investigate?' Structure it - segment, check instrumentation, check seasonality, check a release, isolate the cohort. Also prepare one clean story about a project where you were wrong and caught it.
Why now: Analyst loops are usually four stages: recruiter screen, a timed or live SQL test, a take-home or case, and a stakeholder/behavioural round. The SQL round is a pure filter and most candidates fail it on window functions and multi-table joins, not on syntax.
- practice StrataScratch data analyst SQL and Python questionsStrataScratch · 30 h · Free tier; paid plans for the full question bank
- practice Top SQL 50LeetCode · 25 h · Free
-
8
Target 2 industries, not 'anything'. Rewrite the resume so the top third is SQL, Power BI/Tableau, Python and three named projects with numbers. Apply to 10-15 roles a week, and for every application send one message to someone at the company - alumni, ex-colleagues, local meetup contacts, LinkedIn cold outreach that asks a specific question rather than for a job. Track everything in a sheet.
Why now: In a market where entry-level postings are shrinking and senior roles dominate, the cold-apply conversion rate for a career changer is low single digits. Referrals are the decisive trust signal and roughly reorder the whole funnel. Expect 6-12 months; the people who succeed treat it as a process with a weekly quota, not as an event.
- practice Local and virtual data meetups, Tableau and Power BI user groupsMeetup · 30 h · Free
- practice Maven Analytics Data Playground for portfolio-grade datasetsMaven Analytics · 20 h · Free
Best certifications for a Data Analyst
Which ones matter, what they cost, and how often people pass.
No certification is required to work as a Data Analyst. The one with genuine screening value is Microsoft's PL-300 (Power BI Data Analyst Associate) at $165 for the exam in the United States and 40 to 90 hours of study, and it is worth having if you are targeting Microsoft-shop employers in corporate, healthcare, insurance or government. The Google Data Analytics Professional Certificate costs about $147 to $294 over three to six months and is worth taking as a curriculum, but it carries little weight as a credential on its own: there is no proctored exam and 3.8 million people have enrolled.
Google Data Analytics Professional Certificate
Google (via Coursera)
- Cost
- $49/month; about 6 months at 10 hrs/week, so roughly $150-$300 total, or included in Coursera Plus ($59/mo or $399/yr)
- Study
- About 180 (9 courses, 6 months at 10 hrs/week)
- Pass rate
- No exam; completion-based. Google reports 75% of graduates report a positive career outcome within six months, a self-reported survey figure, not an audited placement rate.
Microsoft Certified: Power BI Data Analyst Associate (Exam PL-300)
Microsoft
- Cost
- $165 exam fee in the US (price varies by country)
- Study
- 60-100 hours of prep for someone who already knows Power BI basics
- Pass rate
- Microsoft does not publish pass rates; passing score is 700/1000, 100 minutes, proctored
Google Advanced Data Analytics Professional Certificate
Google (via Coursera)
- Cost
- $49/month, under 6 months at 10 hrs/week; or included in Coursera Plus
- Study
- About 153 (7 courses)
- Pass rate
- No exam; completion-based
DataCamp Data Analyst Certification (Associate and Professional)
DataCamp
- Cost
- Included with DataCamp Premium (about $28/month billed annually, roughly $330-$336/year; about $35-$39/month billed monthly)
- Study
- Associate: a 2-hour timed exam (DA101) plus a practical exam, with 30 days to complete both
- Pass rate
- Not published; certification valid for two years
Skills employers screen for
Soft skills that decide offers: Translating a vague business question into a measurable one, Writing a two-paragraph summary an executive will actually read, Pushing back on a bad metric without losing the stakeholder, Scoping: knowing when 80% accurate today beats perfect next week, Domain curiosity - learning how the business actually makes money.
Best courses for a Data Analyst
Checked on the provider's page on 16 September 2026. Some links are affiliate links.
We list 6 courses for Data Analyst work, checked on the provider's page. Use one structured course to build the base - the Google Data Analytics Professional Certificate for a complete beginner, or the Power BI path if you already work in spreadsheets - and then spend roughly 120 hours on the three portfolio projects in step 5 of the roadmap. One finished course plus three published projects beats three certificates started.
Browse more Data Analyst courses on Coursera Coursera Plus covers most of the courses above for one monthly fee.
Data Analyst interview questions and format
A Data Analyst loop is typically four stages over two to four weeks: a 25-minute recruiter screen, a 45 to 60 minute SQL assessment rising to window functions, a take-home or live case, then a hiring-manager and stakeholder round. The SQL screen is the stage that filters most candidates, usually on window functions and multi-table joins rather than syntax.
Typically four stages over 2-4 weeks. (1) Recruiter screen, 25 minutes, salary expectations and tool inventory. (2) Technical SQL screen - either a timed online assessment (HackerRank, CoderPad or an internal tool, 45-60 minutes, 3-5 questions rising to window functions) or a live screen-share where you talk through queries. (3) Take-home or case: either a dataset with 3-5 business questions and a 3-8 hour expected effort, or a live case where you verbally investigate a metric drop. Some companies now allow or expect AI assistance on the take-home and will ask you to defend every line. (4) Hiring manager and stakeholder round: behavioural, communication, and 'explain a technical result to a non-technical person'. Excel or BI-tool live exercises appear in finance, insurance and healthcare loops.
Questions that come up
- Write a query returning the second-highest salary per department. (Window functions; expect a follow-up on ties.)
- Given a sessions table and an orders table, compute weekly conversion rate and 4-week rolling retention by signup cohort.
- What is the difference between a LEFT JOIN and an INNER JOIN, and give a case where using the wrong one silently changes a business number.
- Signups dropped 12% month over month. Walk me through how you would investigate.
- We ran an A/B test, the treatment is up 1.4% with p = 0.11. What do you tell the product manager?
- How would you define 'active customer' for this business, and what breaks with your definition?
- Tell me about a time your analysis was wrong. How did you find out and what did you do?
- A stakeholder insists a number in your dashboard is wrong. How do you handle it?
- In Power BI, explain filter context and what CALCULATE actually does.
- How would you check whether a dataset you were just handed is trustworthy?
Prep
Data Analyst FAQ
Can I get hired as a data analyst in 2026 without a degree, if I only have a certificate and portfolio projects?
Yes, but not easily, and not on the certificate alone. What gets career changers interviewed is the combination of fluent SQL (window functions, not just SELECT), three portfolio projects that answer real business questions with a recommendation attached, a domain you already understand from a previous job, and a referral. Degree-free hires are most common at small and mid-size companies, in agency contract roles, and through internal transfer. Large-company graduate programmes usually still screen on the degree, and Indeed's Hiring Lab reported entry-level postings down 7.5 percent year over year in May 2026, so expect the search to take six to twelve months.
Is the Google Data Analytics Professional Certificate worth it in 2026 if I am changing careers from a non-technical job?
Worth it as a curriculum, not as a credential. It is well sequenced, costs about $147 to $294 at $49 a month over three to six months, and gives a beginner a structured 180 hours covering spreadsheets, SQL, R and Tableau. As a signal it is close to worthless on its own: 3.8 million people have enrolled, there is no proctored exam, and Google's figure that 75 percent of graduates report a positive career outcome is a self-reported survey of graduates rather than an audited placement rate. Take it for the structure, then spend the next 120 hours on portfolio projects.
How long does it actually take to become a data analyst if I am working full time in another job?
Plan 9 to 18 months at 10 to 15 hours a week. That splits into roughly six to nine months of skill-building through the 630 roadmap hours on this page, then six to twelve months of searching, because the entry-level rung is the crowded one. If you can move internally at your current employer, it compresses to about six to twelve months in total and you skip the resume screen entirely, which is why internal transfer is the highest-conversion route listed here. Anyone promising three months to a job offer is selling something.
Should I learn Power BI or Tableau to get a data analyst job in 2026?
Pick one and go deep. Power BI if you are targeting corporate, healthcare, insurance, finance, government or anything on a Microsoft stack: it has the larger share of US postings and a well-defined certification in the PL-300, which costs $165 in the United States and takes 40 to 90 hours of study. Tableau if you are targeting technology, media and consumer brands. Learning both shallowly is worse than learning one to the point where you can build a proper star-schema data model and write measures, which is what separates you from the large pool who can only drag fields onto a canvas.
Will AI replace data analysts, and is it still worth learning data analysis in 2026?
It is already replacing part of the job, and it is still worth learning. Large language models write competent SQL, produce first-draft charts and summarise a table faster than you can. What they do not do is know which of four revenue tables is trustworthy, notice that the refunds table double-counts partial refunds, or tell a vice president that the experiment they sponsored did not work. The role is shifting from producing queries to owning metric definitions, experiments and decisions. The US Bureau of Labor Statistics still projects 12 percent growth to 2035 for the operations research analyst code and about 7,500 US openings a year.
Do I need Python to get hired as a data analyst, or is SQL plus Power BI enough?
SQL plus a BI tool is enough for many business intelligence and reporting roles. Python appears in roughly half of Data Analyst postings and becomes close to mandatory at technology companies and for anything labelled product analyst or analytics engineer. Learn SQL to fluency first, because it is the round you get filtered on, then add pandas for the work SQL cannot do: API pulls, reshaping, and automating a recurring report. Do not delay applying until you know Python, and do not list it on a resume until you can use it without a tutorial open.
What salary should I expect for my first data analyst job in the US, and how much does it jump after two years?
Realistically $55,000 to $75,000 depending on city and industry, against a Glassdoor entry-level average of $63,153 and an Indeed junior average of $63,369 as of September 2026. Technology employers pay more: Levels.fyi puts entry-level median total compensation at $80,000, but those roles are the most competitive and the sample is tech-skewed. Expect the real jump at the two to three year mark when you change companies with production experience behind you. That move is usually a 20 to 30 percent step, larger than any internal raise you will be offered.
What does a Data Analyst actually do all day, hour by hour?
Mostly SQL, checking numbers and talking to people. A typical day starts by confirming last night's pipeline run and that the executive dashboard refreshed, then a stand-up with the marketing or product team you sit in. The middle of the day is a long query joining orders, sessions and refunds, and 40 minutes confirming with finance that a table is double-counting. Afternoons bring a dashboard rebuild, a 25-minute ad-hoc request in Slack, an A/B test readout where the honest answer is that the lift is not significant, and documenting a metric definition. Month-end close adds two or three long days.
Is data analyst a good career for someone switching at 40 from an unrelated job?
Yes, and it is one of the better options at 40 specifically because domain knowledge counts. The bar is fluent SQL plus a portfolio, not a computer science degree or calculus, and a background in nursing, accounting, logistics, insurance claims or ad operations is worth more to a hiring manager than another certificate. The honest cost is time and a likely pay cut on entry: median entry pay is around $68,000, against a Glassdoor US average of $93,557 for the role overall. Target the industry you already know, and try the internal-transfer route before applying cold.
Data analyst versus business analyst: what is the actual difference between the two roles?
In most US job postings they are the same skill set pointed at a different question. A Data Analyst is usually measured on producing numbers and dashboards from a database with SQL; a business analyst is more often measured on process, requirements and documentation, and may do less SQL and more stakeholder facilitation. Pay overlaps heavily and titles are used loosely, with marketing analyst, BI analyst, product analyst and revenue operations analyst all covering similar work. Read the tools line in the posting rather than the title: if it names SQL and a BI tool, it is this job.
How much does a data analyst make in New York City compared with the US national median?
About 28 percent more on the offer letter. Our location index puts New York City at 1.28 times the national band and the San Francisco Bay Area at 1.35, so a $68,000 national entry median is roughly $87,000 in New York City and $92,000 in the Bay Area, while Atlanta at 0.98 and Phoenix at 0.95 sit slightly below. The multiplier raises the number on the offer, not what you keep: New York City income tax stacks on top of state tax, and rent absorbs most of the difference. Fully remote analyst roles increasingly pay a national band instead.
Sources
Every number on this page traces to one of these. Page checked 16 September 2026.
- bls.gov/ooh/math/operations-research-analysts.htm
- bls.gov/ooh/math/data-scientists.htm
- bls.gov/news.release/ocwage.t01.htm
- glassdoor.com/Salaries/data-analyst-salary-SRCH_KO0,12.htm
- glassdoor.com/Salaries/entry-level-data-analyst-salary-SRCH_KO0,24.htm
- glassdoor.com/Salaries/senior-data-analyst-salary-SRCH_KO0,19.htm
- indeed.com/career/data-analyst/salaries
- indeed.com/career/junior-data-analyst/salaries
- indeed.com/career/senior-data-analyst/salaries
- levels.fyi/t/data-analyst/locations/united-states
- levels.fyi/t/data-analyst/levels/entry-level/locations/united-states
- levels.fyi/t/data-analyst/levels/senior/locations/united-states
- hiringlab.indeed.com/2026/07/23/the-labor-market-is-tilting-toward-seniority/
- coursera.org/professional-certificates/google-data-analytics
- coursera.org/professional-certificates/google-advanced-data-analytics
- coursera.org/professional-certificates/ibm-data-science
- coursera.org/courseraplus
- learn.microsoft.com/en-us/credentials/certifications/data-analyst-associate/
- datacamp.com/certification/data-analyst
- datacamp.com/tracks/associate-data-analyst-in-sql