How to become a data analyst, without a degree in it

You do not need a degree in data to work as a data analyst. You do, in most cases, need a bachelor's degree in something. O*NET asks its respondents how much education a new hire needs to do the job, and for Business Intelligence Analysts — the occupation sitting closest to what employers post as Data Analyst — 68% answer a bachelor's degree, 23% a master's and 5% an associate's degree (O*NET 15-2051.01, data current to August 2026). The question is about level. No field of study appears anywhere in the answers.

The distance between "a degree" and "a degree in it" is where most career-change advice goes wrong. It treats the subject printed on the diploma as the gate. The federal data points somewhere else: which of four occupations an employer files the job under sets the pay, the tools and the questions in the screen, and the applicant does not get to pick that row. Read a posting for the row it sits in and the rest of the decisions stop being guesses.

Which job is a Data Analyst, in federal data?

There is no BLS occupation called Data Analyst. The work is distributed across four of them, and which one an employer's job is counted under moves the median pay by $41,470.

Occupational Outlook Handbook profiles; wages are May 2025 estimates, employment is 2025.

How to read this. The median is the middle: half the occupation earns more, half less. It is not a starting salary, and for data scientists it is not a Data Analyst salary either — SOC 15-2051 merges Business Intelligence Analysts and Data Scientists into a single code, so the $120,230 median covers machine learning engineers and dashboard builders in the same number. A Data Analyst at an insurance carrier maintaining claims dashboards and a Data Analyst at a consumer brand sizing a market hold the same job title in two different rows of that table.

One row behaves differently from the other three. Statisticians are the only occupation of the four where BLS assigns a master's degree as the typical entry-level education, and the only one where the degree requirement in a posting is likely to be both real and specific about a subject. It is also the rarest of these jobs by a wide margin: 31,300 of the 1,372,700 people counted across all four rows, against 952,700 market research analysts. A career changer reading "Statistician" and "Data Analyst" as interchangeable titles is reading one gate as though it were the other.

What degree does a data analyst job ask for?

Education O*NET's respondents say a new hire needs

O*NET asks "how much education does a new hire need to perform a job in this occupation?" Data current to August 2026. Only the three most common answers are published per occupation, so no column totals 100%.

	Business Intelligence Analysts	Data Scientists	Market Research Analysts
Bachelor's degree	68%	48%	57%
Master's degree	23%	44%	39%
Associate's degree	5%	4%	—
Post-baccalaureate certificate	—	—	4%

Read the second row before the first, because the second row is where the money goes. Nearly half of the respondents on the data scientist code — 44% — say a new hire needs a master's, and 39% say the same for market research analysts. That is a real number and pretending otherwise would be dishonest. Two things sit against it. The first is in the same figure: on Business Intelligence Analysts, the code closest to a posted Data Analyst job, 68% say a bachelor's is enough. The second is that BLS, which assigns entry-level education by a different method entirely, gives "bachelor's degree" to three of the four occupations in the table above, data scientists included.

Act on the cheaper reading first and apply with the bachelor's. A master's bought before the first application is two years of tuition spent against a requirement that may not apply to the row you land in, and the figure above cannot tell you which row that will be.

What this data does not tell you: nobody was asked what the degree should be in. O*NET's question is about level only, and the answers come from workers in the occupation and from occupational experts rather than from a count of live postings — informed judgement about a typical new hire, not a tally of what employers wrote down last month. Whether a history degree clears the bar is exactly the question the survey does not ask.

Do you need experience to get the first one?

Two federal sources disagree here, and the disagreement is the useful part. BLS lists, for data scientists, "Work experience in a related occupation: None" and "On-the-job training: None." O*NET files the same work in Job Zone Four and says employees "usually need several years of work-related experience, on-the-job training, and/or vocational training." Both are accurate about different things: BLS is describing the minimum to enter the occupation, O*NET is describing what the average person already in it has accumulated. When a Data Analyst posting asks for three to five years, it is quoting the second one.

What does the bottom of the range pay?

10th percentile annual wage, May 2025

The wage below which the lowest-paid tenth of each occupation falls. A Data Analyst posting is counted under one of these four.

Data scientists		$67,240
Statisticians		$64,000
Operations research analysts		$57,060
Market research analysts		$43,390

Medians describe people in the middle of a career, so someone moving into a first Data Analyst job should read the chart above instead. The spread between its ends is $23,850, and almost none of that is explained by how good the analyst is. It is explained by who hired them. BLS reports median wages for data scientists by industry in the same May 2025 release: $142,240 in publishing, broadcasting and content providers, $132,380 in computer systems design, $129,490 in credit intermediation, $128,050 in management of companies and enterprises, and $108,650 in insurance carriers. The same dashboard work, built for a streaming service and built for an insurer, is separated by $33,590 at the median before a word of negotiation.

What the 10th percentile is not: an entry-level wage. It is the bottom decile of everyone in the occupation, which includes part-year workers, low-cost metros, and people who have held the role for a decade without moving. A first Data Analyst job in a mid-size city lands somewhere near that figure rather than exactly on it, and federal data publishes no first-job wage for any occupation.

Which tools show up in the job?

O*NET keeps a technology list for each occupation, grouped by software category. Those categories are printed alphabetically, so the sequence carries no information — the signal is which tools O*NET flags as hot technologies, a flag it derives from counting employer job postings. For Business Intelligence Analysts the flagged tools include:

• Business intelligence and data analysis software — Alteryx, Apache Spark, Google Looker Analytics, Microsoft Power BI
• Database management system software — Amazon DynamoDB, Apache Hive, Elasticsearch, Oracle PL/SQL
• Database user interface and query software — IBM DB2, Transact-SQL, Amazon EC2
• Data mining software — Google Analytics, Snowflake
• Spreadsheet and presentation software — Microsoft Excel and Microsoft PowerPoint, the two here that O*NET additionally marks in demand

PowerPoint carrying that in-demand mark is not a joke entry, and neither is Excel outranking Snowflake on it. Set the list beside the flagged tools for Data Scientists — IBM SPSS Statistics, SAS, TensorFlow and MATLAB under analytical software, PyTorch under database query tools — and two different jobs come into focus wearing similar titles. Market Research Analysts sit elsewhere again: Tableau, Google Looker, Google Analytics, IBM SPSS and Minitab. The top-rated task for Business Intelligence Analysts, at an importance score of 91, is "Generate standard or custom reports summarizing business, financial, or economic data." Reporting is the job. Modeling is the other job.

What the flag is not: a measure of what anyone does all day. O*NET counted Python in 814,960 of the 46.9 million US job postings it scanned in 2025 — how often employers type the word, not how many hours it is open on a screen. It is still the closest public read on demand, and it has a blunt consequence for a career changer. Six months on a deep learning course buys less than six weeks spent getting genuinely fast with SQL and one BI tool, because the first is preparation for the occupation you are not applying to. O*NET also lists a registered apprenticeship under Business Intelligence Analysts — Business Intelligence Engineer — worth knowing about, since apprenticeship is the one entry route that pays while it teaches.

Does a certificate count instead of a degree?

Partly, and the honest version of the answer is an uncomfortable one. On Market Research Analysts and Marketing Specialists, 4% of O*NET's respondents say a post-baccalaureate certificate is what a new hire needs — a real answer, but one that sits after a bachelor's degree rather than in place of it. For Business Intelligence Analysts the only sub-bachelor's answer common enough to be published is the 5% who say an associate's degree. "No degree" appears in neither list of published answers, which is not the same as nobody saying it; it means fewer said it than said associate's.

Where a certificate does real work is the second half of the Job Zone Four description. The requirement O*NET records is a four-year degree plus several years of related experience, and a certificate argues for the second half far better than the first. Take a bookkeeper who finishes a SQL and Power BI course, then rebuilds the monthly close reporting at their own employer and applies for the internal Data Analyst opening six months later: the course is not the credential doing the work, the rebuilt report is. The same bookkeeper applying to an unknown company with the certificate and nothing built has evidence for neither half of what the occupation asks.

The four steps, when your degree is in something else

Getting to a first Data Analyst application

Ordered by what the O*NET and BLS occupation profiles say the work actually requires.

1. 1
   Pick the row, not the title

   Decide whether you are aiming at market research analyst work (customer and industry data, Excel and Tableau, 952,700 jobs in 2025) or business intelligence work (SQL, Power BI, dashboards). They pay $78,760 and $120,230 at the May 2025 median and they expect different portfolios.

2. 2
   Learn SQL and one BI tool properly

   O*NET flags Transact-SQL, Microsoft Power BI, Alteryx and Apache Spark as hot technologies for Business Intelligence Analysts, and marks Excel in demand. One tool learned to the point of fluency beats four learned to the point of a certificate.

3. 3
   Build two reports, not ten dashboards

   The occupation's highest-rated task is generating reports that summarize business, financial or economic data, scored 91 for importance. A portfolio that shows a question being asked and answered matches that description. A gallery of charts does not.

4. 4
   Apply to the adjacent titles too

   O*NET's related occupations for Business Intelligence Analysts are Clinical Data Managers, Computer Systems Analysts, Data Scientists, Financial Risk Specialists, Market Research Analysts and Operations Research Analysts. Searching only for 'Data Analyst' hides most of the openings.

The order matters more than the contents. Choosing the target occupation first is what makes every later choice obvious, and skipping that step is why so many career-change portfolios contain a neural network notebook and no SQL at all.

Where are the openings?

BLS projects data scientist employment to grow 35% between 2025 and 2035, an increase of 95,400 jobs, with about 24,800 openings a year on average once transfers out of the occupation and retirements are counted. Market research analysts grow more slowly in percentage terms — 7%, or 66,300 jobs — but from a base of 952,700 people, more than three times the 275,600 counted as data scientists.

Growth rate and opening count are not the same thing, and for a first job the second one is what matters. On the same 2025–35 projections, BLS expects about 82,000 openings a year for market research analysts against about 24,800 for data scientists — 3.3 times the flow into the slower-growing occupation. O*NET's separate 2024–34 vintage puts the gap wider still, at 87,200 projected annual openings for Market Research Analysts and Marketing Specialists against 23,400 for the data scientist code that also holds Business Intelligence Analysts. Operations research analysts are smaller than both, at 12% growth, 13,500 additional jobs and about 7,500 openings a year. So the fastest-growing of these occupations is not the one with the most doors in it, and a career changer chasing the 35% headline is walking toward the narrower entrance.

A projection is not a vacancy count. None of these figures say what share of those openings will go to somebody with no prior analyst title, and that is the number nobody publishes.

What else is this job called?

Search one title and you see a fraction of the market. O*NET's related occupations for Business Intelligence Analysts are Clinical Data Managers, Computer Systems Analysts, Data Scientists, Financial Risk Specialists, Market Research Analysts and Marketing Specialists, and Operations Research Analysts. From the market research side, the related list points back at Business Intelligence Analysts, Data Scientists, Management Analysts, and Financial and Investment Analysts, alongside Marketing Managers and Search Marketing Strategists. Employers post the same work as Reporting Analyst, BI Analyst and Insights Analyst without using the word data in the title at all.

If you are applying for this role

Three things, specifically.

Search the title variants together rather than one at a time — Data Analyst, BI Analyst, Reporting Analyst, Business Intelligence Analyst, Market Research Analyst — because they are the same work filed under different SOC codes by different recruiters. Expect a first offer somewhere between the 10th percentile and the median of whichever occupation the employer sits in: $43,390 to $78,760 for market research work, $67,240 to $120,230 for business intelligence work, on May 2025 estimates. And ask one question in the screen, early: what does the person in this seat produce every week? A weekly report answers to the Business Intelligence Analyst description. An ad hoc model answers to a different one, and the answer tells you which pay row you are standing in before you talk about money.

Applying across five title variants means retyping the same education and work history into five different forms. HeroApply fills those forms from your resume, so the breadth costs you time you have rather than time you do not.