Skip to content
APEX FLOWACADEMY
MENU
Technology / beginner

Data Analytics From Scratch: Answer Real Business Questions

Ask a sharp question, clean messy data, explore it, compare fairly, chart it honestly and write a recommendation a boss can act on, using a spreadsheet, SQL or both.

24 lessons6 modulesabout 8 hours9 workbook itemsbeginner
READ LESSON 1 FREE

One payment of $39. Instant online access to the full written course. No subscription. Refund policy.

What you will be able to do

You collect, clean, analyse and present data so a boss can act on it. By the end you will have worked through the whole analytics loop on one practice dataset and produced a case study with charts and a written recommendation that states how sure you are and what the data cannot show.

Who it is for. Complete beginners who want to answer questions with data: office workers, small-business owners, students and career changers. You do not need coding or maths beyond adding, counting, dividing and comparing.

You finish with. Corner Kettle Case Study: A Portfolio Case Study With a Written Recommendation. Using the synthetic Corner Kettle data, answer your Question Card from start to finish. Deliver one project folder that another person can rerun and one answer-first memo. The folder shows the question, the source and its fingerprint, the cleaning, the exploration, the analysis with its limits, three honest charts, and a check that your key numbers match the fixed answers. You may extend the work to a real dataset you are allowed to use, but the practice data is the version that is checked.

Certificate. Finish every lesson, resolve every quiz question with at least 50% right on the first try, and tick the capstone checklist — Apex Flow Academy issues a verifiable Certificate of Completion with a unique ID and a public verification page. It is a certificate of completion, not a degree, licence, accreditation or exam result.

The path, module by module24 lessons
01 · Ask a Question and Meet Your Data4 lessons

See the whole analytics loop in miniature, turn a vague request into a sharp question with a decision rule, create the practice dataset for Corner Kettle (a made-up shop), and learn to read a table before you touch it.

  1. 1.1The Six-Stage Loop and Your Practice ShopFREE
  2. 1.2Turn a Vague Request Into a Question Data Can Answer
  3. 1.3Create Your Practice Data and Open It
  4. 1.4Read a Table: Rows, Columns and the Data Dictionary
02 · Get the Data and Clean It4 lessons

Learn where data comes from and how to protect the raw file, audit a messy table with a fixed set of checks, fix text, date and number problems, and decide what to do with missing, repeated and impossible rows while keeping a cleaning log.

  1. 2.1Know Your Source and Protect the Raw File
  2. 2.2Audit the Data Before You Fix Anything
  3. 2.3Fix Text, Date and Number Problems
  4. 2.4Handle Missing, Repeated and Impossible Rows, and Keep a Log
03 · Explore: Totals, Groups, Trends and Odd Values4 lessons

Get to know the clean data before you draw conclusions: summarize with count, sum, mean and median, split into groups, follow the months, and find and handle unusual values. You finish with a set of exploration notes.

  1. 3.1Summarize With Count, Sum, Mean and Median
  2. 3.2Group and Compare With GROUP BY and Pivot Tables
  3. 3.3Follow the Months: Trends and Seasons
  4. 3.4Spread and Outliers: Find Unusual Values and Decide What to Do
04 · Analyze: Compare Fairly and Test How Sure You Are4 lessons

Turn exploration into an answer: compare groups with rates, join a second table safely, avoid mistaking correlation for cause, and check how much your result depends on small counts and on your own cleaning choices.

  1. 4.1Compare Groups With Rates, Not Raw Counts
  2. 4.2Join a Second Table Without Losing or Doubling Rows
  3. 4.3Correlation Is Not Cause: Ask What Else Changed
  4. 4.4How Sure Are You? Small Counts and Stress Tests
05 · Chart It Clearly and Read Charts Honestly4 lessons

Choose the chart that fits the question, build a clear bar chart in a spreadsheet and from code, learn how charts and averages mislead so you can avoid it, and check every chart against a short honesty list.

  1. 5.1Match the Chart to the Question
  2. 5.2Build a Clear Bar Chart: Spreadsheet and Code
  3. 5.3Charts That Lie: Axes, Windows and Pictures
  4. 5.4Averages and Percentages: Always Ask 'Of What?'
06 · Explain It and Finish Your Case Study4 lessons

Turn your results into a one-page answer-first memo, write findings a boss can act on, package the project so anyone can rerun it and check it, and finish with the capstone: a portfolio case study with a written recommendation.

  1. 6.1Know Your Reader and Lead With the Answer
  2. 6.2Write Findings a Boss Can Act On
  3. 6.3Package a Repeatable Case Study With an Answer Check
  4. 6.4Capstone: Rebuild, Review and Present Your Case Study
Free preview · lesson 1.1No sign-up
Free lesson · 1.1

The Six-Stage Loop and Your Practice Shop

About 5 minStep 1 of 7 · You will be able toNo sign-up
Step 1 of 71 min read

You will be able to

  • Name the six stages of the analytics loop, in order.
  • Say what you will build in this course.
  • Choose your tools: a spreadsheet, SQL, or both.
Step 2 of 71 min read

Why this matters

Data analytics means turning rows of facts into a clear answer that someone can act on. Beginners often start with a chart. That is backwards. A chart built on a fuzzy question or messy data looks neat and still misleads. A fixed order of work protects you from that.

Step 3 of 72 min read

Learn it

The loop has six stages. Each module of this course teaches one.

  1. Ask. Write the question and the decision it supports (Module 1).
  2. Get and clean. Collect the data, check it, and fix the mess (Module 2).
  3. Explore. Look at totals, groups, trends and odd values (Module 3).
  4. Analyze. Compare fairly and test how sure you can be (Module 4).
  5. Chart. Show the answer so it is hard to misread (Module 5).
  6. Explain. Tell the reader what to do and why (Module 6).

It is a loop because the last stage often raises a new question. It is also a loop because a later stage can send you back. If you find mess while exploring, you return to cleaning.

Your practice project. You will study Corner Kettle, a made-up online shop that sells tea, mugs, gifts and gear. The owner has one year of orders and one worry: where should the shop put its effort next year? All the data is synthetic. That means a script creates it, and no real person or business is in it. The numbers say nothing about real shops.

Your tools. Pick a route, or use both.

  • Spreadsheet route. Google Sheets works with a Google account, so check its current terms. LibreOffice Calc is free. Excel usually needs a paid plan, so check its current options.
  • SQL route. SQL is a language for asking questions of tables. You will use SQLite, which comes inside Python. It needs no server and costs nothing.

Python 3 is free from python.org. You use it only to run scripts we give you. They build the data, clean it and draw charts. You do not need to write Python from scratch.

Every number printed in this course came from running the code beside it. Your results should match. When they match, you know you are on track.

Step 4 of 71 min read

See it in action

Example (illustrative numbers). Six orders arrive with messy spelling. Here is the whole loop in a few lines of Python. Ask: which channel sold the most? Get: the six orders. Clean: strip spaces and use lower case, so 'Search' and 'search' become one group. Explore and analyze: add up each channel and its share. Chart: bars made of # signs.

Pythonyou can edit this
Output
search   105  54%  #####################
email     55  28%  ###########
social    35  18%  #######

Explain: search sold the most in this small sample, 105 of 195 (54%), so it gets the first look. Without the cleaning line, 'email' and 'EMAIL' would count as two channels, and the answer would change.

If this is not working: you do not need to run this example yet, because lesson 3 sets up Python. If you try it and see an error, check that you copied every line.

Step 5 of 71 min read

Common mistakes

  • Starting with the chart. It feels fast, but it hides a weak question. Write the question first.
  • Skipping the clean step because the data 'looks fine'. Spaces and capital letters are hard to see. Always check the values.
  • Treating the loop as one-way. Real work goes back and forth. Expect it.
  • Thinking you need advanced maths. This course uses adding, counting, dividing and comparing.
Step 6 of 71 min read

You are done when

You can list the six stages from memory, and you know which route you will use. A folder named corner-kettle exists on your computer.

Step 7 of 7

You finished the free lesson

That is one lesson from the course. The full course gives you every remaining lesson, a quick check and a hands-on task in each one, and the workbook of templates and checklists.

Full course$39

What you need
  • A computer running Windows, macOS or Linux, and a web browser
  • A spreadsheet program: Google Sheets (used with a Google account; check its current terms), LibreOffice Calc (free) or Microsoft Excel (usually a paid plan; check its current options). The course works with any of them, and you can choose the SQL route instead
  • Python 3.10 or newer, free from python.org. It runs the scripts we give you and includes SQLite, the database used for the SQL route, so nothing else is needed
  • A plain-text editor such as Notepad, TextEdit in plain-text mode, or a free code editor
  • Optional: DB Browser for SQLite, a free open-source tool for running SQL without the terminal (check its current features)
  • All core lessons can be completed for free. The practice data is synthetic and is created by a script in Module 1
Before you start
  • You can save and find a file on your computer
  • You can open a spreadsheet program and type in a cell
  • You are willing to type a short command in a terminal; the course shows you how
The capstone

Corner Kettle Case Study: A Portfolio Case Study With a Written Recommendation

Using the synthetic Corner Kettle data, answer your Question Card from start to finish. Deliver one project folder that another person can rerun and one answer-first memo. The folder shows the question, the source and its fingerprint, the cleaning, the exploration, the analysis with its limits, three honest charts, and a check that your key numbers match the fixed answers. You may extend the work to a real dataset you are allowed to use, but the practice data is the version that is checked.

  • Question Card with decision, metric, guardrail, comparison, scope and a decision rule written before you looked at results (Module 1)
  • Data dictionary with grain, key and definitions of revenue and net sales (Module 1)
  • Source Note with fingerprint, audit list and cleaning log (408 rows in, 397 kept, 11 set aside) (Module 2)
  • Exploration notes with summary numbers, channel table, monthly shape and your decision on the bulk orders (Module 3)
  • Findings draft with return rates, join checks, careful wording on the April discount, rough ranges and a stress test (Module 4)
  • Chart pack of three charts with finding titles, units, notes and honesty checks (Module 5)
  • One-page answer-first memo with a written recommendation, plus a README with the run order and check_answers.py showing 8 of 8 (Module 6)
Key terms taught30
Data analytics
Turning rows of facts into a clear answer that someone can act on.
Analytics loop
The six stages of the work: ask, get and clean, explore, analyze, chart and explain. Later stages can send you back to earlier ones.
Synthetic data
Made-up data created by a script. It describes no real person or business.
Seed
The starting number for a random-number maker. The same seed gives the same 'random' results every time.
CSV file
A plain text file where each line is a row and commas separate the columns.
Grain
What one row of a table stands for, such as one order for one product.
Key
A column that should be different on every row, such as order_id, or that links two tables, such as customer_id.
Data dictionary
A list of every column with its type and meaning, plus the definitions of your key numbers.
Raw data
The original file, exactly as received. You keep it untouched and work on copies.
Fingerprint (hash)
A short code worked out from a file. It changes if even one character changes, so it proves the file is the same.
Personal data
Anything that identifies a person, such as a name, email address, phone number or exact home address.
Audit
A fixed set of checks that gives a counted list of problems in a dataset before you fix anything.

Browse the full Academy encyclopedia

How this course was checked

22 checks were run and recorded while writing this course (code, formulas, commands and facts), and it lists 22 official sources it was checked against. Prices, features and policies of outside tools can change, so check each tool's own website.

Created by Apex Flow Academy with AI assistance. For education only; not legal, tax, financial or medical advice. Results depend on your effort and circumstances.