Financial Management and Business Data Analytics · Data Processing, Organisation, Cleaning and Validation

Data Organisation and Data Structures in Business Data Analytics

Updated 10 October 2026 · Fact-checked

Data organisation is arranging raw data in a logical hierarchy: characters form fields, fields form records, records form files, and files form a database. Data is then classed as structured (fixed rows and columns), semi-structured (tagged, flexible) or unstructured (no fixed format). To answer questions, identify the level or the type, then justify with an example.

Understand Data Organisation and Data Structures

Raw data is useful only when it is arranged so that you can find, update and analyse it. Data organisation is the way data is arranged and stored. Think of a customer register in a shop. It has a neat layout, and every entry follows the same pattern.

The usual hierarchy goes from smallest to largest. A character is a single letter, digit or symbol. A field is a group of characters that gives one fact, such as customer name or PAN. A record is a set of related fields about one entity, such as everything about one customer. A file is a collection of similar records, such as the customer file. A database is an organised collection of related files that can be stored, searched and shared. Some books add a bit and a byte below the character. Know the order and give a business example for each level.

A key field (primary key) identifies each record uniquely, for example an invoice number or employee ID. Two records should never share the same primary key. A key in another file that points back to it is a foreign key. This is how files in a relational database are linked.

Data is also classified by form. Structured data follows a fixed schema of rows and columns, such as ledger entries, sales tables and bank statements in a spreadsheet or SQL database. It is easy to sort, filter and total. Unstructured data has no predefined format, such as emails, images, audio, video, social media posts and scanned documents. It needs special tools to analyse. Semi-structured data has no rigid table layout but carries tags or markers that give some structure. Examples are XML, JSON files and HTML pages.

For exams, remember that most business data in accounting systems is structured, while a large share of data generated today is unstructured. The type of data decides the storage and the analysis tool you can use.

Key rules to remember

Data hierarchy: Character → Field → Record → File → Database; Each level is built from the one before it. Reverse the order when a question asks you to break a database down.
Field vs record: Record = set of related fields about one entity; A field is one attribute (e.g. Invoice Date). A record is the full row for one invoice.
Primary key rule: Primary key = unique, non-repeating identifier for each record; Used to link and retrieve records. A foreign key in another file refers to a primary key.
Types of data by structure: Structured | Semi-structured | Unstructured; Structured: fixed rows and columns. Semi-structured: tags, e.g. JSON, XML. Unstructured: no format, e.g. images, emails.

How to solve Data Organisation and Data Structures questions

Use this method for definition, identification, difference and classification questions on data organisation.

1Read the question and decide whether it asks about the hierarchy (field, record, file, database) or about the type of data (structured, semi-structured, unstructured).
2For hierarchy questions, find the unit being described. One fact is a field. One full row about one entity is a record. Many similar records form a file. Linked files form a database.
3For type questions, ask whether the data fits fixed rows and columns. If yes, it is structured. If it has tags but a flexible layout, it is semi-structured. If it has no set format, it is unstructured.
4Name the key field if the question involves identifying or linking records.
5Support each answer with a business example from the case, such as an invoice, payroll or customer file.
6For difference questions, compare on at least three points: format, storage, ease of analysis and examples.
7Close with a one-line conclusion on why the organisation matters, such as faster retrieval, less duplication or better analysis.

Quickest way: Smallest-to-largest and fixed-format test

When to use it: Use this for MCQs and short identification questions where you have under a minute.

For hierarchy, ask: is it one fact (field), one entity's full set of facts (record), a collection of the same type of records (file), or many linked files (database)?
For type, ask: can I put it in a fixed table? Yes means structured.
If not in a table but it has tags or key-value labels, pick semi-structured.
If it is free text, image, audio or video, pick unstructured.
Eliminate options that swap two levels, such as calling a row a field.

Common mistakes in Data Organisation and Data Structures

Calling a whole row a field.
Students see a cell and a row as the same thing in a spreadsheet.
Fix: A column heading or single cell value is the field. The whole row about one entity is the record.
Treating a file as the same as a database.
Both words suggest stored data, so they seem interchangeable.
Fix: A file holds similar records of one kind. A database holds several related files that are linked.
Classing emails or scanned invoices as structured because they are used in accounting.
Students confuse the business use of data with its format.
Fix: Judge by format. Free text and images are unstructured, even if they relate to accounts. The extracted entries in a ledger table are structured.
Saying JSON or XML is unstructured.
They are not tables, so students assume there is no structure.
Fix: They carry tags or keys that give structure, so they are semi-structured.
Writing differences without examples.
Students memorise definitions only.
Fix: Add one business example for each type or level. Examples earn marks and prove understanding.
Ignoring the key field in record-related answers.
The hierarchy is learned but the unique identifier is forgotten.
Fix: Mention that each record has a unique primary key, such as employee ID or invoice number, whenever you discuss records or linking files.

Worked examples

Example 1

A company stores payroll data. For employee E102, it holds Name: Ravi Sharma, Department: Finance, Basic Pay: ₹45,000. All such employee entries are kept together in one payroll listing. The HR, Payroll and Attendance listings are linked through Employee ID. Identify the field, record, file and database, and the key field.

Show the solution

Name, Department and Basic Pay are single facts, so each is a field.
All fields for employee E102 together form one record.
The collection of all employee records in the payroll listing is the payroll file.
HR, Payroll and Attendance files linked together form the database.
Employee ID identifies each employee uniquely and links the files, so it is the primary key.

Answer: Field: Name, Department or Basic Pay. Record: all details of E102. File: the payroll listing. Database: the linked HR, Payroll and Attendance files. Key field: Employee ID.

Example 2

Classify the following as structured, semi-structured or unstructured, with reasons: (a) a sales table of invoice number, date and amount in a SQL database; (b) a JSON file of customer orders received from a website; (c) recorded customer complaint calls.

Show the solution

(a) The sales table has fixed rows and columns with defined data types. It is structured.
(b) The JSON file has no table layout but uses keys and tags to label values. It is semi-structured.
(c) Audio recordings have no predefined format or fields. They are unstructured.
Note the effect on analysis: (a) can be queried directly, (b) needs parsing, and (c) needs conversion such as speech-to-text before analysis.

Answer: (a) Structured, (b) semi-structured, (c) unstructured, each justified by the presence or absence of a fixed schema or tags.

Exam tips

Learn the hierarchy in order and attach one business example to each level. Examiners often test it as an MCQ with options in a jumbled order.
For the difference between structured and unstructured data, write at least three comparison points and one example for each.
In case-based questions, identify the format first, not the business use. This avoids classification errors.
Mention the primary key whenever the question talks about records, linking or avoiding duplication.
There is no negative marking in the MCQ section, so attempt every question. Use the elimination approach if unsure.

Practice questions from Data Processing, Organisation, Cleaning and Validation

Data Organisation and Data Structures in other exams

The same ground in other exams, if you are preparing for more than one or want another angle on it.

Data Organisation and Data Structures: frequently asked questions

What is the difference between a field and a record?

A field is a single data item, such as a customer name. A record is a set of related fields about one entity, such as all details of that customer. In a table, columns are fields and rows are records.

What is the difference between structured and unstructured data?

Structured data fits a fixed format of rows and columns, like a sales ledger table. Unstructured data has no set format, like emails, images or videos. Structured data is easier to search and total, while unstructured data needs special tools.

Is a file the same as a database?

No. A file is a collection of similar records, such as a supplier file. A database is an organised collection of related files that can be linked and accessed together.

Where does semi-structured data fit?

It sits between the two. It has no rigid table but uses tags or keys to label the data. JSON, XML and HTML are common examples.