Lesson 1 of 16 · Level 1: Beginner
What Is Data Analytics and What Does a Data Analyst Do?
- ⏱ 14 min read
- ❓ 4 review questions
- ✍️ Exercise with solution
- 📥 Downloadable template
In this lesson
- What data analytics is, in plain English, and how it turns raw records into business decisions.
- The four types of analytics (descriptive, diagnostic, predictive and prescriptive) and the question each one answers.
- What a data analyst actually does all day, and how the job differs from a data scientist or a data engineer.
- The core toolkit you'll learn in this course: spreadsheets, SQL, a BI tool and Python.
- What the U.S. Bureau of Labor Statistics says about pay and growth in data-related jobs.
Every lesson in this course uses the same practice file: 200 fictional orders from a U.S. retailer in 2025. Download it now and open the Sales Clean sheet. You don't need to do anything with it yet; just scroll through and get familiar with the columns.
- sales-practice-data.xlsx · Excel or Google Sheets
Data analytics is the practice of collecting, cleaning and examining data to answer a business question and help someone make a better decision. A data analyst is the person who does that work: they take a messy question ("Why were sales so good in December?"), find the right data, check that it's trustworthy, calculate what matters and explain the answer in a way a manager can act on. In this lesson you'll see what that looks like with real numbers from the course dataset, which tools analysts use, and whether it's a realistic career to aim for.
What is data analytics?
Every business generates records: a sale at a register, a click on a website, a support ticket, a shipment. One record by itself says almost nothing. Thousands of them, organized and summarized, can tell you which products sell, which customers leave and which campaigns pay for themselves. Data analytics is the bridge between those records and a decision.
A useful way to picture it is a four-step ladder:
- Data: raw facts. "Order SO-10001, January 1, 2025, Pennsylvania, 2 office chairs at $249."
- Information: data organized so a person can read it. "The West region brought in $36,626 in 2025."
- Insight: information that explains something. "Most of the West's revenue came from 14 large wholesale orders, not from everyday shoppers."
- Decision: what the business does about it. "Assign a dedicated account manager to wholesale buyers in the West."
Analysts spend their days moving up that ladder. The higher you climb, the more valuable your work is to the people who make decisions.
The four types of analytics
Not every question is the same kind of question. Analysts usually group them into four types, from the simplest to the most ambitious:
Descriptive: what happened?
Totals, averages, counts and trends. "Revenue in 2025 was $89,735 from 200 orders."
Where you'll learn it: spreadsheets, descriptive statistics and pivot tables (Lessons 4, 6 and 7).Diagnostic: why did it happen?
Breaking a number down to find the cause. "December hit $18,319 because six wholesale orders added $14,961."
Where you'll learn it: SQL, pivot tables and correlation (Lessons 7, 9 and 13).Predictive: what is likely to happen?
Trends and forecasts. "If the pattern holds, Q1 2026 will be the slowest quarter again."
Where you'll learn it: trends and moving averages (Lesson 13).Prescriptive: what should we do?
Recommendations and tests. "Run an A/B test before doubling the holiday ad budget."
Where you'll learn it: storytelling, A/B testing and the portfolio project (Lessons 12, 13 and 15).Most entry-level analyst work is descriptive and diagnostic. That's not a downside: a clear, correct answer to "what happened and why?" is exactly what most managers are missing. Predictive and prescriptive work builds on that foundation, so you need the first two to be solid before the others mean anything.
1. A store manager asks: "Why did online sales drop in October compared with September?" What type of analytics is that?
The manager already knows what happened (sales dropped). They want to know why, which is diagnostic analytics: breaking the number down by product, region or channel to find the cause.
A first look at the practice data
Here's a quick descriptive answer from the course dataset to show what analysis produces. The question: where does our revenue come from? Grouping the 200 orders by U.S. Census region gives this split:
| Region | Revenue | Orders | Average order |
|---|---|---|---|
| West | $36,626 | 64 | $572.28 |
| Northeast | $20,786 | 45 | $461.91 |
| South | $18,684 | 59 | $316.68 |
| Midwest | $13,639 | 32 | $426.22 |
| Total | $89,735 | 200 | $448.68 |
Notice what a good analyst does next: they don't stop at "the West wins." The South has 59 orders, almost as many as the West's 64, yet it brings in about half the revenue. Why? Is it smaller orders, cheaper products, or fewer business customers? That curiosity, turning a descriptive number into a diagnostic question, is the core habit of the job. You'll answer that exact question in Lesson 13.
What does a data analyst do all day?
Job titles vary (data analyst, business analyst, reporting analyst, BI analyst, operations analyst), but the work follows a similar rhythm. Here's a realistic day for an analyst at a mid-size retailer:
- 9:00 a.m.Clarify a requestThe sales director asks for "a quick look at wholesale." A 15-minute call turns that into a precise question: which regions and products drive wholesale revenue, and is it growing?
- 10:00 a.m.Pull the dataA SQL query extracts 2025 orders from the company database. Product costs come from a second table.
- 11:00 a.m.Clean and checkRemove duplicate rows, fix state names typed three different ways and confirm the totals match the finance report.
- 1:30 p.m.AnalyzeA pivot table shows wholesale is 18.5% of orders but 68.5% of revenue. That's the headline.
- 4:00 p.m.ShareUpdate the dashboard and send a five-line summary with one chart and a recommendation.
Typical deliverables include recurring reports (weekly sales, monthly KPIs), dashboards in tools like Power BI or Looker Studio, one-off analyses that answer a specific question, and data quality checks. Communication is a bigger part of the job than most beginners expect: an analysis nobody understands doesn't change anything.
Data analyst vs. data scientist vs. data engineer
These roles overlap, and small companies often blend them. The general split looks like this:
- SQL and spreadsheets
- ●●●●
- Statistics and ML
- ●●●●
- Typical output
- Reports, dashboards, recommendations.
- SQL and spreadsheets
- ●●●●
- Statistics and ML
- ●●●●
- Typical output
- Forecasts, predictive models, experiments.
- SQL and spreadsheets
- ●●●●
- Statistics and ML
- ●●●●
- Typical output
- Databases, data pipelines, automation.
Neither path is "better"; they suit different interests, and many people move between them.
2. Which task is most typical of a data analyst rather than a data engineer?
Analysts use data that already exists to answer business questions and communicate results. Building and maintaining pipelines and infrastructure is data engineering work.
The data analyst toolkit
You don't need dozens of tools. A small, solid stack covers most analyst jobs in the U.S., and this course teaches each piece in order:
| Tool | What it's for | Where you'll learn it |
|---|---|---|
| Excel or Google Sheets | Quick analysis, formulas, pivot tables, small datasets | Lesson 4 and Lesson 7 |
| SQL | Pulling and summarizing data from databases | Lesson 8 and Lesson 9 |
| Power BI or Looker Studio | Interactive dashboards that update automatically | Lesson 11 |
| Python (pandas) | Repeatable cleaning and analysis, larger files | Lesson 14 |
| Statistics | Knowing when a number is meaningful and when it's noise | Lesson 6 and Lesson 13 |
Beyond tools, four non-technical skills separate good analysts from average ones: curiosity (asking "why?" one more time), business sense (knowing which numbers matter to the company), healthy skepticism (checking whether a surprising number is real or a data error) and clear communication (one chart and three sentences beat a 40-slide deck).
Is data analytics a good career?
The U.S. Bureau of Labor Statistics (BLS) doesn't track "data analyst" as a single occupation. Analyst jobs are spread across several categories. Here are three related ones from the BLS Occupational Outlook Handbook, with median pay from May 2025 wage data and projected employment growth for 2025–2035:
| BLS occupation | Median annual pay (May 2025) | Projected growth, 2025–35 |
|---|---|---|
| Data scientists | $120,230 | 35% |
| Operations research analysts | $88,940 | 12% |
| Market research analysts | $78,760 | 7% |
A median is the middle value across all experience levels, so entry-level offers are usually lower, and pay varies a lot by city, industry and company size. BLS describes data scientists' projected growth as much faster than the average for all occupations, but a growing field still has competition for junior roles. A strong portfolio (Lesson 15) is how you stand out. Source: BLS Occupational Outlook Handbook, figures checked in November 2026.
3. According to the BLS figures above, what was the median annual pay for operations research analysts in May 2025?
Operations research analysts had a median of $88,940. $120,230 is the figure for data scientists and $78,760 for market research analysts. Always check which occupation a salary statistic refers to before quoting it.
Common myths about data analytics
"You need to be a math genius."
Most analyst work uses arithmetic, percentages, averages and basic statistics. Logic and clear thinking matter more than advanced math.
"You must learn to code before anything else."
Start with spreadsheets and SQL. Python is powerful, but many analysts use it only part of the time.
"More data always means better answers."
A clean, relevant dataset beats a huge messy one. Garbage in, garbage out.
"The analysis speaks for itself."
It never does. You have to explain what the numbers mean and what to do about them.
How this course is organized
The 16 lessons are grouped into four levels. Each one builds on the previous, and all of them use the same dataset, so by the end you'll know these 200 orders better than the fictional company's own sales team.
- Beginner: process, data types and spreadsheets
- Intermediate: cleaning, statistics, pivots and SQL basics
- Advanced: joins, charts, dashboards and storytelling
- Expert: correlation, A/B tests, pandas, portfolio and job search
4. In the practice data, the South has 59 orders and the West has 64, but the West earns almost twice the revenue. What is the best next step for an analyst?
A gap like this is a diagnostic question, not a conclusion. Jumping to "Western customers are richer" is a guess. Breaking the number down is how you find the real driver (spoiler: it's mostly wholesale orders).
Practice exercise
Classify each question as descriptive, diagnostic, predictive or prescriptive. Then open the Summary sheet of the practice file and find one number that answers question (a).
- (a) How many orders did we receive in 2025?
- (b) Why did April revenue jump to $15,221 when March was $3,117?
- (c) How much revenue should we expect in January 2026?
- (d) Should we offer free shipping to wholesale customers in the Midwest?
- (e) What share of revenue came from the Online channel?
- (f) Are customers who buy smart watches more likely to buy earbuds later?
See solution
- (a) Descriptive. The Summary sheet shows 200 orders (formula
=COUNTA('Sales Clean'!A2:A201)). - (b) Diagnostic. It asks for the cause of a change. (Two large wholesale smart watch orders in the West, $4,776 and $4,378, explain most of it.)
- (c) Predictive. It asks about the future.
- (d) Prescriptive. It asks what to do. You'd answer it with analysis plus a test, not with a single number.
- (e) Descriptive. Online brought in $15,861, which is 17.7% of $89,735.
- (f) This one is a trick: it's diagnostic or predictive in spirit, but you can't answer it with this dataset because there's no customer ID column. Knowing what your data can't answer is part of the job.
Summary
- Data analytics turns raw records into information, insights and, finally, decisions.
- The four types of analytics answer: what happened, why, what's likely next and what we should do.
- Analysts clarify questions, pull and clean data, analyze it and explain the results; communication is half the job.
- The core toolkit is spreadsheets, SQL, a BI tool and Python, plus solid basic statistics.
- BLS doesn't track "data analyst" as one occupation; related roles show strong projected growth, but entry-level competition is real.
Next lesson: The Data Analysis Process in 6 Phases. You'll follow one real question about December's sales spike from a vague request to a recommendation.
Create your account in 30 seconds, mark your lessons and download your certificate at the end.