← All entries
Data and AnalyticsEntry 13

DataAnalyst

Three different jobs share this title. One builds reports, one answers questions, one moves data around. The posting decides which one you get.

The same work is also posted as
Business Intelligence AnalystReporting AnalystAnalytics AnalystInsights AnalystData and Reporting SpecialistBusiness Analyst

What the title covers

What stays constant is that you work with the company's data and produce something another person acts on. What changes is which part of that you spend your time doing. Reporting work builds and maintains dashboards so people can see numbers without having to ask. Analysis work takes one question and answers it, which means locating the data, deciding whether it can be trusted, and explaining what it does and does not show. Pipeline work builds the routes that move data out of the systems that create it and into somewhere it can be queried. Most postings mix two of the three. In every version, a large share of the day is spent discovering that the data is wrong.

A second use of this title

Same title, three different jobs

The tools named in a posting tell you which version it is, more reliably than the description does. SQL alongside Tableau, Power BI or Looker describes reporting work. Python or R alongside statistics, experiments or testing describes analysis work. Airflow, dbt, warehouses or the letters ETL describe pipeline work. A posting listing all of them is either a small team where one person does everything or a company that has not decided what it wants.

The difference matters because the three lead to different places. Reporting leads toward business intelligence and eventually toward running an analytics team. Statistical analysis leads toward data science. Pipeline work leads toward data engineering, which pays the most of the three and is the hardest to move into later if you did not start there.

Where it is found

Every industry now employs them, which is why the title has stretched to cover so much.

Technology companiesThe most defined versions of the role, usually with the three jobs split across separate people and clear titles for each.
Banks, insurers and healthcareLarge volumes of regulated data, more rules about access, and slower but steadier work.
Retail and consumer brandsHeavy on sales, inventory and customer behaviour, with visible seasonal cycles.
Government, agencies and consultanciesAnalysis performed for someone else, often with less control over how the data was collected.

Typical entry requirements

Education
Bachelor's degree. Statistics, economics, mathematics, business and computer science are common, though people arrive from many majors
Expected
SQL. This appears in nearly every posting regardless of which version of the job it is
Frequently required
A dashboard tool such as Tableau, Power BI or Looker
Often listed
Python or R, most seriously at employers where the work leans toward statistics
Experience
Zero to two years. A portfolio built on real public data carries unusual weight here, sometimes more than the degree
Example of a taskAnalytics

What a typical day may look like

The request below is typical of how this kind of work is assigned. The steps that follow describe what completing it involves.

Your analytics manager9:15 AM
Sign ups dropped eighteen percent last week. Before anyone panics, can you tell me whether that is real?

Answer by end of day.

01

Check whether the tracking changed before checking whether behaviour changed. A release, an altered tag or a broken event explains most sudden drops.

02

Confirm you are comparing like with like, including whether a holiday, the end of a campaign or a partial week is involved.

03

Break the drop down by segment rather than looking at the total, because eighteen percent overall is usually one channel falling much further.

04

Check the figure against a source collected a different way, such as the product database rather than the analytics tool.

05

Decide whether the change is larger than the normal movement between weeks, rather than treating every decline as a signal.

06

Write the answer as one sentence at the top with the chart underneath, so it can be forwarded without you there to explain it.

07

Say what you would need in order to be certain and how long that would take, instead of implying more confidence than the data supports.

What is delivered

A short written answer stating whether the drop is real, one chart showing it by segment, and a note listing what you ruled out. Usually posted where the question was asked rather than sent as a document.

FitData Analyst

Is this job for you

Characteristics of the work itself, listed so you can compare them against how you prefer to spend your time.

Yes, if you like

  • Questions that have a findable answer
  • Being the person who checks before everyone else reacts
  • Working alone for long stretches
  • Cleaning something messy until it can be trusted
  • Explaining a number to someone who will act on it
  • Being right mattering more than being quick

No, if you don't like

  • Deciding what to do about what you found
  • Data that arrives clean and complete
  • Credit when the number confirms what everyone hoped
  • A steady supply of interesting questions rather than routine requests
  • Building the product rather than measuring it
  • Answers that are certain rather than probable
CompensationUnited States

What this work pays

Two numbers, because they answer two different questions. The first is what you would be paid in your first job. The second is what people in this field earn after they have been doing it for years.

Starting out
$68,873Average starting salary, business graduates, Class of 2026

This is the average across all business graduates rather than this role specifically. Analytics roles usually start at or above it, and the more technical versions start higher still.

After several years
$91,290Middle of the range across everyone in this field

Government data has no category called data analyst, so this figure uses operations research analysts, the closest published match for business facing analytical work.

The government files this title in three different places

Depending on what the job actually involves, this work is counted as market research analysis, where the middle of the range is around seventy seven thousand, as operations research analysis at roughly ninety one thousand, or as data science at about a hundred and thirteen thousand. That spread is not noise. It is the same three jobs described further up this page, showing up in the wage data. Which band you land in is predicted better by the tools listed in the posting than by anything in the title.

SOURCES / NATIONAL ASSOCIATION OF COLLEGES AND EMPLOYERS, WINTER 2026 SALARY SURVEY. U.S. BUREAU OF LABOR STATISTICS, OCCUPATIONAL OUTLOOK HANDBOOK, OPERATIONS RESEARCH ANALYSTS, MARKET RESEARCH ANALYSTS AND DATA SCIENTISTS, MAY 2024 WAGE DATA. GOVERNMENT DATA DOES NOT LIST DATA ANALYST SEPARATELY.

Reading a posting against your own background

SIGNAL scores a specific job posting against your resume and shows you where you match and where you do not.

Try SIGNAL free