Your First Look at Data
Data comes in three flavours — structured, semi-structured, and unstructured. Understanding the difference is the first step to working with data in the cloud.
What is data?
Data is just information written down so a computer can use it.
Think of a supermarket. The price sticker on a can of beans is data. The photo on the label is data. The handwritten note from your flatmate saying “we need more beans” is also data.
All three are information — but they’re organised very differently. That difference matters because it changes how computers store, search, and analyse them.
The three types of data
Structured data
Structured data fits neatly into rows and columns — like a spreadsheet or a database table. Every row follows the same pattern, and every column has a defined type (text, number, date).
Aisha’s example: The campus food ordering app stores orders in a table:
| OrderID | Student | Item | Price | Date |
|---|---|---|---|---|
| 1001 | Aisha | Chicken wrap | 8.50 | 2026-04-20 |
| 1002 | Liam | Coffee | 4.00 | 2026-04-20 |
Every order has the same fields. You can sort by price, filter by date, or count orders per student — because the structure is predictable.
Key features:
- Fixed schema (the column names and types are defined in advance)
- Stored in relational databases or tabular files (CSV, Excel)
- Easy to query using languages like SQL
- Works best when every record has the same fields
Semi-structured data
Semi-structured data has some organisation, but it’s flexible. Not every record needs the same fields. The data describes itself using tags, keys, or markers.
Aisha’s example: Her app also stores user profiles, but some students fill in more details than others:
// Student 1 — minimal profile
"name": "Aisha", "email": "aisha@uni.ac.nz"
// Student 2 — detailed profile
"name": "Liam", "email": "liam@uni.ac.nz",
"phone": "021-555-0199", "dietary": "vegetarian"Liam has extra fields that Aisha doesn’t. That’s fine — the data is flexible. Each record carries its own labels (keys like “name”, “dietary”), so the system knows what each value means.
Key features:
- Flexible schema — records can have different fields
- Self-describing — tags or keys label each value
- Common formats: JSON, XML, YAML
- Stored in document databases (like Azure Cosmos DB) or NoSQL stores
Unstructured data
Unstructured data has no predefined format. It’s the photos, videos, audio files, PDFs, and emails that don’t fit into rows or columns.
Aisha’s example: Students upload photos of their meals for reviews. Each photo is a blob of pixels — no columns, no keys, no rows. A computer can store it and display it, but it can’t “query” the photo the way it queries a table.
Key features:
- No schema — no rows, columns, or tags
- Examples: images, videos, audio, PDFs, Word documents
- Stored in file storage or blob storage (like Azure Blob Storage)
- Requires specialised tools (AI, search engines) to analyse content
| Feature | Structured | Semi-Structured | Unstructured |
|---|---|---|---|
| Schema | Fixed (defined upfront) | Flexible (self-describing) | None |
| Format | Tables (rows & columns) | JSON, XML, YAML | Images, video, audio, PDFs |
| Queryable? | Yes — SQL queries | Yes — with the right tools | Not directly — needs AI or search |
| Example | Order table in a database | User profile in JSON | Photo uploaded by a customer |
| Azure storage | Azure SQL Database | Azure Cosmos DB | Azure Blob Storage |
Exam tip: how the exam tests this
The exam loves giving you a scenario and asking “what type of data is this?” Look for these clues:
- Fixed columns, every row the same → Structured
- Tags or keys, flexible fields → Semi-structured
- Files with no internal organisation → Unstructured
A tricky pattern: “an email” can be semi-structured (it has To, From, Subject fields) OR unstructured (the body text). Similarly, log files can be semi-structured (if they have timestamps and severity levels in a consistent format) or unstructured (if they’re free-form text). The exam usually clarifies by asking about a specific aspect.
Why does it matter?
The type of data determines:
- Where you store it — relational database? Document database? Blob storage?
- How you query it — SQL? API calls? AI-powered search?
- How you process it — direct queries? ETL pipelines? Machine learning?
Choosing the wrong storage for your data type leads to wasted money, slow performance, or lost information. The rest of this course teaches you how to make those choices on Azure.
Real-world mix: most systems use all three
Real applications almost always combine all three types. Take Priya’s FreshMart grocery chain:
- Structured: Sales transactions (item, price, quantity, date, store)
- Semi-structured: Product metadata from suppliers (JSON files with varying attributes per category)
- Unstructured: Security camera footage, customer feedback emails, scanned invoices
A complete data solution handles all three — that’s why Azure offers different services for each.
Meet the characters
Throughout this course, you’ll follow four people working with data:
| Character | Who They Are | Data Scenarios |
|---|---|---|
| 📈 Priya Sharma | Senior Data Analyst at FreshMart Groceries (50 stores) | Analysing sales trends, building dashboards, Power BI reports |
| 💼 Tom Chen | Business Ops Manager at Pacific Freight (logistics, 200 drivers) | Tracking deliveries, querying databases, optimising routes |
| 🎓 Aisha Mohammed | CS student at Auckland University (building a campus food app) | Learning data basics, choosing storage, first database decisions |
| 🏢 Jake Bennett | Solo DBA at CloudPulse (SaaS startup, 15 developers) | Managing production databases, scaling storage, choosing Azure services |
Flashcards
Knowledge check
Aisha's food ordering app stores meal reviews as free-text comments with no fixed format. What type of data is this?
Tom's logistics company receives shipment data from partners in JSON format. Each partner includes different fields — some include weight, others include dimensions. What type of data is this?
Which of the following is the BEST description of structured data?
Next up: Data File Formats: CSV, JSON, Parquet & More — learn how data is physically stored in files.