Choose a Data Storage Approach at the Cloud

Introduction

Not all data is the same, and not all storage works the same way.

Just like you wouldn't store clothes, food, and important papers in the same place at home, businesses also need different storage solutions for different types of data.

Let's take an online shopping website as an example. It deals with many kinds of data:

Product information (like names, prices, and descriptions)
Photos and videos of the products
Financial data (like customer orders, payments, and profits)

Each of these types of data is different and used in different ways. Some need to be loaded very fast, some need to be super secure, and some take up a lot of space.

So, it's important to:

Understand what kind of data you have
Know how that data will be used
Choose the best storage method to keep your app fast and efficient

There's no one-size-fits-all solution, so picking the right storage for each type of data helps your business run better.

Classify your data

An online shopping business deals with many types of information. And each type of information may need a different way to store it.

To pick the best way to store your data, you first need to understand what kind of data you have . We usually group data into three types:

1. Structured Data

This is neat and organized—like data in a spreadsheet or a table.

Example: Product prices, names, stock levels, and customer orders.

This kind of data is easy to search, sort, and analyze.

2. Semi-Structured Data

This data is somewhat organized, but not as neat as structured data.

Example: Customer reviews or website logs (like who clicked what).

It has some structure, but not enough to fit into a simple table.

3. Unstructured Data

This is messy data with no clear structure.

Example: Product images, videos, or customer support chat recordings.

You can't put this kind of data into rows and columns.

By knowing which type your data is, you can choose the best storage option that keeps things running smoothly and quickly for your business.

Approaches to storing data in the cloud

Structured data

Structured data is information that is neatly organized—like data in a spreadsheet.

Every piece of data follows the same format and has the same kind of information in it.

In structured data, sometimes called relational data , all data has the same fields or properties. All the data has the same organization and shape, or schema . The shared schema allows this type of data to be easily searched by using query languages like Structured Query Language (SQL). This capability makes this data style perfect for applications like CRM systems, reservations, and inventory management.

For example, imagine a table that stores student names. Every row has:

A student ID
A student name
An age

All rows look the same. That's what makes it "structured."

Structured data is often stored in database tables with rows and columns. In the table, a key column indicates how one row in a table relates to data in another row of another table. In the following image, a table that has data about grades gets data from a table of student names and a table of class data by using key columns.

Why Use Structured Data?

Because it's so organized, it's easy to:

Search
Sort
Filter
Connect with other data

That's why structured data is perfect for things like:

Customer databases (CRM)
Hotel or flight reservations
Inventory systems
Example: Tables with Structured Data
Let's say we're running a school system. We might have three tables :

Why Relationships Matter

We don't repeat all the student or class information in the Grades table.
Instead, we use IDs to link data between tables.
This keeps data clean, saves space, and is easy to manage.

One Thing to Remember

Since everything has to follow the same format, making changes is harder .

If you want to add a new field (like student phone number), you must update every record to match the new structure.

Semi-structured data

Semi-structured data is a type of data that is partly organized , but not as strictly as data in a spreadsheet or database table. It doesn't always fit neatly into rows and columns, but it still has some structure to make it understandable.

For example, instead of a table, the data might look like this:

  
    {
  "name": "John",
  "age": 30,
  "city": "New York"
}

This kind of data is often called NoSQL or non-relational data because it's not stored in traditional databases.

Real-Life Examples of Semi-Structured Data:

Emails : The subject, sender, and message are labeled, but not stored in rows and columns.
JSON files : Used by websites and apps to send data.
XML files : Common in older systems to share data between apps.
Log files : Data from software or websites that isn't strictly organized but has patterns.

To share or store this kind of data, developers use something called a data serialization language . Serialization is just the process of turning data into a format that can be saved to a file or sent over the internet.

What is Serialization?

Serialization is like packing your data into a suitcase so you can send it to someone else or store it somewhere.

You take the data in your computer's memory and turn it into a format (like JSON or XML) that can be easily saved or shared.
When another system gets that data, it can unpack it (called deserialization) and use it, even if it's a completely different system.

This is helpful because two different systems (like two computers or apps) don't need to know much about each other. As long as they both understand the same data format (like JSON or XML), they can read and use the data without confusion.

In simple words

Semi-structured data = "Kinda organized" data that uses tags, labels, or keys, but not strict tables.
Serialization = "Packing data" into a common format so it's easy to share or save.
Why it's useful : Two different apps or computers can understand each other's data as long as they use the same data format.

Common serialization languages

Three common serialization languages are XML, JSON, and YAML.

Let's understand one by one all three of these.

XML

XML (Extensible Markup Language) is one of the oldest and most popular ways to store and share data.

It's text-based , which means you can open it in a text editor and understand it easily.
It's also machine-readable , so computers can read and process it quickly.
Almost every programming language has tools (called parsers) to read XML.

Why Use XML?

You can organize data clearly with labels (called tags).
You can show relationships between pieces of data (like a person and their hobbies).
XML has rules for schemas (to define structure), transformations (to change data format), and even for displaying on websites .

Example: A Person's Details in XML

  
    <Person Age="23">
    <FirstName>Quinn</FirstName>
    <LastName>Anderson</LastName>
    <Hobbies>
        <Hobby Type="Sports">Golf</Hobby>
        <Hobby Type="Leisure">Reading</Hobby>
        <Hobby Type="Leisure">Guitar</Hobby>
   </Hobbies>
</Person>

Here's what's happening:

<Person> is the main tag , and Age="23" is an attribute (extra info about the person).
<FirstName> and <LastName> are elements that store data.
<Hobbies> contains child elements <Hobby> to show a list of hobbies

Easy Way to Think About It:

XML is like putting your data in labeled boxes :

The labels (tags) tell you what's inside.
Boxes can be placed inside other boxes (child elements).
Attributes are like sticky notes on the box giving extra details.

Pros and Cons

✅ Easy to read by humans and computers.

✅ Very flexible; can describe complex data.

❌ Can be wordy and take up more space , which makes it slower to send over the internet.

Because of this, simpler formats like JSON have become more popular for web apps.

JSON

JSON (JavaScript Object Notation) is a very popular way to store and share data, especially on the web.

It's lightweight , meaning it doesn't take up much space.
It uses curly braces { } to organize data.
It's easier to read and write compared to XML because it's less wordy .
JSON is often used by websites and apps to send data between servers and browsers.

Example: A Person's Details in JSON

  
    {
    "firstName": "Quinn",
    "lastName": "Anderson",
    "age": "23",
    "hobbies": [
        { 
           "type": "Sports",
           "value": "Golf" 
        },
        {
           "type": "Leisure",
           "value": "Reading" 
        },
        { 
           "type": "Leisure", 
           "value": "Guitar" 
        }
    ]
}

Here's what's happening:

Each piece of data is written as a key/value pair , like "firstName": "Quinn" .
Curly braces { } hold related information together.
Square brackets [ ] represent a list of items (like hobbies).

Easy Way to Think About It

JSON is like writing data as a shopping list with labels :

You write the label (key) and its value next to it.
Lists (like hobbies) are grouped together inside brackets.
It's simple, clean, and easy for programmers to understand.

Why Developers Love JSON

✅ Smaller and faster to send over the internet.

✅ Very easy to use with JavaScript (and most programming languages).

✅ Great for web apps and mobile apps.

❌ Not as formal as XML, so it doesn't enforce strict rules.

❌ Can be harder for non-programmers to edit because of symbols like {}, [], and :.

🔑 In short: JSON is a lightweight, modern way to store and share data, while XML is more structured but wordy . JSON is now the preferred choice for most web services .

YAML

YAML (YAML Ain't Markup Language) is a newer way to store and share data.

It's designed to be super easy for humans to read and write .
Instead of using lots of symbols like {} , [] , or <> , YAML uses indentation and line breaks to organize data.
Because it's so clean and simple, YAML is often used for configuration files —files that people write to tell a program how to run.

Example: A Person's Details in YAML

  
    firstName: Quinn
lastName: Anderson
age: 23
hobbies:
  - type: Sports
    value: Golf
  - type: Leisure
    value: Reading
  - type: Leisure
    value: Guitar

Here's what's happening:

Each piece of data is written as key: value (like firstName: Quinn ).
Indentation (spaces at the start of a line) shows relationships, instead of brackets or tags.
The - symbol is used for lists (like hobbies).

Easy Way to Think About It

YAML is like writing notes in plain English :

No need for extra punctuation or brackets.
Just use spaces and dashes to keep things organized.
Much simpler to read at a glance than XML or JSON.

Why Developers Use YAML

✅ Very human-friendly —easy to read and write.
✅ Cleaner than JSON or XML (no curly braces or angle brackets).
✅ Perfect for config files (like Docker, Kubernetes, GitHub Actions).
❌ Requires careful indentation ; a wrong space can break things.
❌ Not as widely supported as JSON yet, but still growing in popularity.

🔑 In short: YAML is the cleanest and easiest-to-read format, great for configuration files and data that people often write. It's like a simpler, prettier version of JSON.

Here's a simple side-by-side comparison table of XML, JSON, and YAML in a way that's easy to remember:

Feature	XML 🏷️	JSON 📦	YAML 📝
Full Name	Extensible Markup Language	JavaScript Object Notation	YAML Ain't Markup Language
Look	Uses <tags> and attributes	Uses {}, [], :	Uses indentation and -
Example	<name>Quinn</name>	"name": "Quinn"	name: Quinn
Readability	Harder (lots of symbols)	Easier than XML	Easiest (clean, minimal symbols)
File Size	Larger (verbose)	Smaller	Smallest
Structure	Strict, very formal	Less strict, key/value pairs	Very flexible, indentation-based
Best For	Documents, web services (older)	APIs, web apps, mobile apps	Config files, DevOps tools
Pros	Widely supported, very structured	Lightweight, fast, popular	Most human-friendly, simple
Cons	Verbose, harder to read	Symbols can confuse non-techies	Indentation errors break things
Common Use Cases	RSS feeds, SOAP, legacy systems	REST APIs, web data exchange	Docker, Kubernetes, GitHub Actions, configs

🔑 Summary

XML = Very structured but heavy. Good for older systems and when you need strict formatting.
JSON = Lightweight and easy for apps. Most popular for APIs and web data today.
YAML = Cleanest and easiest to read. Best for writing configuration files.

Unstructured Data

Unstructured data is data that has no fixed structure or organization . It doesn't fit neatly into rows, columns, or tables like a spreadsheet.

Often, it comes as files (like photos, videos, or documents).
Even though a video file or image might have some basic info about it (like the date it was taken or file size), the actual content (the image pixels or video frames) is not organized in a way computers can easily put into a database.

Easy Way to Think About It

Unstructured data is like a box of random stuff :

You know what's in the box (a photo, a video, a document), but the computer can't easily "read" and organize the content.
It takes extra tools (like AI or special software) to make sense of it.

Examples of Unstructured Data

🎥 Media files : Photos, videos, music, voice recordings.
📄 Microsoft 365 files : Word documents, PowerPoints.
📝 Text files : Notes, articles, ebooks.
🖥️ Log files : Records created by systems or apps, but not neatly organized.

Key Points

✅ Great for storing rich information (images, videos, etc.).
❌ Harder for computers to organize or search through without extra processing.

🔑 In short
Unstructured data is messy but valuable —it's all your photos, videos, and documents that don't naturally fit into a database table.

Here's a clear and simple comparison table of Structured, Semi-Structured, and Unstructured Data :

Feature	Structured Data 📊	Semi-Structured Data 📂	Unstructured Data 🎥
Definition	Fully organized and stored in fixed rows & columns.	Partly organized with tags or labels, but not in strict tables.	No fixed format or structure; hard to organize.
How It Looks	Neat spreadsheets or databases.	Tagged data like JSON, XML.	Files like photos, videos, and documents.
Examples	Bank records, sales data, and inventory tables.	JSON, XML, CSV, NoSQL data.	Images, videos, Word docs, emails, logs.
Storage	Relational databases (SQL).	NoSQL databases, data lakes.	File systems, cloud storage.
Ease of Search	🔍 Super easy (SQL queries).	🔍 Easy with some effort.	🔍 Hard; needs AI or special tools.
Human Readability	Low (mostly numbers and IDs).	Medium (tags help).	High (photos, videos, plain text).
Best For	Reports, analytics, structured apps.	Web services, APIs, flexible data sharing.	Storing rich media, documents, and logs.
Key Takeaway	Perfectly organized data.	"Kinda organized" with tags.	Messy, raw, and free-form.

🔑 Summary

Structured = Organized like a clean spreadsheet.
Semi-structured = Has some structure (tags/labels) but not strict.
Unstructured = Messy/raw data like videos, images, and documents.

Data Classification: Understanding Your Data Types

We can sort data into three main types :

Structured – Very organized, fits perfectly into tables and columns.
Semi-structured – Somewhat organized, has labels or tags, but not everything is the same.
Unstructured – Not organized, doesn't fit into a table, often raw files like images or videos.

Knowing which type of data you have helps you choose the right way to store and manage it .

🔍 Examples From an Online Retail Business

🛍️ Product Catalog Data → Semi-Structured

Your product catalog includes details like product ID, price, size, colors, photos, and videos.
At first, everything looks structured because all products have the same kind of information.
But as you add new product features , like "Bluetooth-enabled" shoes, you don't want to edit every single product to add that field.
This means your data is not perfectly uniform anymore.
Because of these differences, your product catalog becomes semi-structured —it's organized with tags or fields, but not every product has the same fields.

🎥 Photos and Videos → Unstructured

Product images and videos are unstructured because they're just media files.
Even though they may have metadata (like file size or date taken), the actual image or video content doesn't fit neatly into rows and columns.

📊 Business Data → Structured

Data like sales records, inventory levels, and monthly performance is structured .
This is because you want to compare numbers over time , run reports, and analyze trends.
For that, the data needs to be in organized tables with consistent formats.

📝 Summary

Data Type	Example from Retail Business	Why It's Classified This Way
Structured	Sales reports, inventory numbers	Data is neat, organized, and easy to analyze.
Semi-structured	Product catalog (with optional fields like Bluetooth)	Data is mostly organized, but different products have different fields.
Unstructured	Product images and videos	No clear table format; content is raw and not uniform.

🔑 In plain words:

Structured data is like a perfectly organized spreadsheet .
Semi-structured data is like a flexible list where some items have extra details.
Unstructured data is like a photo album or video collection —lots of useful information, but not organized for quick computer analysis.

Introduction

Classify your data

1. Structured Data

2. Semi-Structured Data

3. Unstructured Data

Approaches to storing data in the cloud

Structured data

Why Use Structured Data?

Why Relationships Matter

One Thing to Remember

Semi-structured data

What is Serialization?

Common serialization languages

Pros and Cons

JSON

Why Developers Love JSON

YAML

Conclusion