Descriptive Analysis: Key Types, Questions, and Insights from Sales Data
Updated: Jan 11
Introduction
Descriptive analysis helps us understand what has happened in the data by summarizing and organizing information in a meaningful way. In this blog, we focus on descriptive analysis using a cleaned sales dataset. We begin by explaining what descriptive analysis is and why it is important in data analytics. Next, we discuss the key types of descriptive analysis in simple terms. We then frame important descriptive questions based on sales data. Using these questions, we generate clear and easy-to-understand insights. The goal is to highlight patterns, trends, and distributions in sales performance. This approach helps transform raw sales data into useful business understanding.
What Is Descriptive Analysis?
Descriptive analysis is the process of using basic statistical techniques to summarize and describe a dataset. It is popular because it helps turn raw data into clear and easy-to-understand insights.
Descriptive analysis does not predict future outcomes like other data analysis techniques. Instead, it focuses on past data and organizes it in a way that makes the information clear and meaningful.
Why is Descriptive analysis important?
The analysis is important because it helps us understand what has already happened in the data. It summarizes large datasets into simple numbers and visuals that are easy to interpret. This makes it easier to identify patterns, trends, and unusual values. Descriptive analysis provides a clear foundation before moving to advanced analysis. It supports better reporting and informed decision-making.
There are four key types of descriptive analysis
1. Measures of Central Tendency
2. Measures of Dispersion (Variability)
3. Measures of Frequency / Distribution
4. Measures of Position
1. Measures of Central Tendency
Measures of central tendency describe the typical or average value in a dataset. They help summarize large amounts of data into a single meaningful number. Common measures include mean, median, and mode.
For example: In sales data, calculating the average sales amount per order helps understand the typical value of a sale.
2. Measures of Dispersion (Variability)
Measures of dispersion explain how spread out the data values are. They show whether the values are close together or widely different. Common measures include range, variance, and standard deviation.
For example: Analyzing the variation in sales amounts helps understand whether sales are consistent or fluctuate significantly across orders.
3. Measures of Frequency / Distribution
Frequency and distribution measures show how often values occur within a dataset. They help identify patterns, trends, and dominant categories.
For example: Counting the number of sales orders by region helps determine which region has the highest sales activity.
4. Measures of Position
Measures of position describe where a value stands relative to others in the dataset. They use percentiles and quartiles to identify relative ranking.
For example: Finding the top 25% of sales amounts helps identify high-value transactions in the sales data.
Descriptive Analysis Visualizations Using Tableau
After cleaning the sales dataset using Python in my previous blog, the data was imported into Tableau to perform descriptive analysis. Tableau was chosen because it allows quick exploration of data through visualizations and makes insights easy to understand for both technical and non-technical audiences.
If you would like to know how to clean the data set using python, please refer to my blog
The cleaned sales data set looks like below.
What is the average sales amount per order?
This KPI shows the typical sales value per order using the mean. It is useful for a quick performance summary and is easy for readers to understand. If there are a few very high or very low orders, the average can shift, so it is best presented along with the median.
What is the median sales amount?
Median represents the middle sales value when all orders are sorted. It is important because it is less affected by outliers than the average.
The median sales amount is $290, which is lower than the average, indicating that most orders are of lower value while a few high-value transactions increase the overall average.
How much do sales amounts vary across orders?
A box plot summarizes variability using:
Median (center line)
Quartiles (box)
Whiskers (spread)
Outliers (dots)
This chart quickly shows whether sales amounts are tightly clustered or widely spread and whether there are extreme orders.
Sales amounts show high variability across orders, with most transactions clustered in a moderate range and a few high-value orders acting as outliers that increase overall spread.
How many orders were placed per region?
Orders were highest in the East region, followed by North and West. A small number of orders fall under “Not Reported”, indicating missing region information. These records were retained to ensure data completeness and transparency.
What percentage of total orders comes from each region?
The East region dominates order volume with 36.84% of total orders, while North and West contribute similar shares. A small portion of orders (5.26%) lacks region information and is retained as “Not Reported” to maintain data completeness.
What is the Sales percentage per Region?
The East region contributes the largest share of total sales revenue at 53.25%. South and West show moderate contributions, while North records a negative sales share due to net returns exceeding purchases. This highlights the importance of handling returns separately when analyzing revenue distribution.
What sales amount falls in the top 25% of orders?
The sales amount that marks the top 25% of orders is $925.00, meaning only one-quarter of orders exceed this value.
Which orders are considered high-value transactions?
High-value transactions are defined as orders with sales amounts at or above the 75th percentile ($925). The chart displays individual orders, clearly highlighting those high-value transactions while also showing lower-value and return orders for context.
Conclusion
This blog demonstrated how descriptive analysis can be used to understand sales data after proper data cleaning. By using a cleaned sales dataset, we ensured that the insights were accurate and reliable. The analysis revealed typical order values, variation in sales amounts, and differences in regional performance. It also identified the sales threshold for top-performing orders and highlighted high-value transactions. Overall, descriptive analysis helped convert sales data into clear and meaningful insights that support better business understanding.
