ADIDAS Sales Analysis
Unraveling the Secrets of Adidas’ Sales Success: A Data-Driven Analysis
Introduction:
In the fiercely competitive world of sports apparel, understanding market trends, consumer preferences, and sales dynamics is crucial for staying ahead. Adidas, a global leader in this industry, is no stranger to these challenges. In this in-depth analysis, we dive into a comprehensive dataset to uncover the hidden patterns and strategies behind Adidas’ sales triumphs.
Our exploration reveals fascinating insights into several key aspects: from sales performance and product popularity to regional market variations and efficiency of sales methods. We delve into questions such as which products are leading in sales, which regions show robust performance, and the interplay between profit margins and sales figures.
This article presents our findings from an exhaustive exploratory data analysis (EDA), offering a rare glimpse into Adidas’ business strategies. Whether you’re a marketing professional, a data enthusiast, or simply a fan of the brand, join us in uncovering the secrets behind Adidas’ success story, one data point at a time.
Project Overview:
This project aims to provide a thorough analysis of Adidas’ sales performance and market presence. Utilizing a detailed dataset, the project explores various facets of the business, including sales trends, product popularity, regional market performance, and more.
Requirements:
For this project, we utilized the “Adidas Sales Dataset” from Kaggle, in XLSX format. Analysis was performed using VS Code Python environment, complemented by Tableau for creating interactive dashboards. The dataset, organized in an Excel spreadsheet, includes the following columns:
- Retailer: Names such as Foot Locker, Walmart, Sports Direct, and West Gear.
- Retailer ID: A unique identifier for each retailer.
- Invoice Date: The date the invoice was generated.
- Region: Geographic areas like West, Northeast, Southeast, South, and Mid-west.
- Product: Categories like Men’s and Women’s Street and Athletic Footwear, and Apparel.
- Price per Unit: The cost of a single item of a specific product.
- Units Sold: The quantity of a product sold in a specific timeframe.
- Total Sales: The total revenue from sales over a certain period.
- Operating Profit: A measure of the profitability from the main business operations.
- Operating Margin: A ratio indicating the profitability of the business operations.
- Sales Method: The channels through which sales were made, including In-store, Outlet, and Online.
The first step towards any data analysis project/ task should be Exploratory Data Analysis (EDA).
Exploratory Data Analysis (EDA):
Exploratory Data Analysis (EDA) serves as a critical phase in the data analysis process, offering an in-depth look into the dataset’s characteristics and uncovering pivotal insights. In this stage, we thoroughly examine the data to spot trends, anomalies, and relationships that inform our subsequent analysis. EDA enables us to detect missing data, outliers, and key correlations while identifying significant variables. Through various visualization techniques, we gain insights into the data’s distribution and central tendencies, unraveling its underlying patterns. This process is fundamental in shaping hypotheses and directing the focus of further detailed analysis. The EDA phase allowed us to:
- Gain a clear understanding of the dataset’s structure, including its size and key characteristics.
- Spot and remove any duplicate entries to ensure the accuracy of our findings.
- Chart the progression of sales over different timeframes, notably years and months.
- Investigate how sales figures varied month-to-month.
This step was important and crucial in establishing a solid base for informed decision-making, enabling us to extract valuable insights from the data with greater precision and relevance.
Business problems that can be solved using this dataset:
1. Sales Performance Analysis: Which products are excelling in sales, and which are underperforming?
2. Regional Market Analysis: Which stores are experiencing strong sales, and which ones are lagging?
3. Profit Margin Analysis: Does the profit margin significantly impact sales?
4. Efficiency of Sales Methods: Which sales method is more effective — in-store or online?
5. Price Optimization: Is there a specific price range that achieves better sales than others?
6. Product Portfolio Optimization I: Determine which products are most profitable, segmented by location.
7. Market Expansion Opportunities: Assess the best and worst performing stores based on their locations.
8. Time Series Analysis: Investigate whether there has been a consistent sales trend over time or any noticeable monthly trends.
9. Predictive Sales Analysis: Develop a forecast for monthly sales.
10. Tableau Sales Dashboard
Project:
Data Cleaning and Pre-Processing:
Data Cleaning and Pre-Processing are pivotal steps to ensure the integrity and usability of almost every dataset. I meticulously sifted through the data, correcting inconsistencies, handling missing values, and standardizing formats. This foundational work was essential for accurate analysis and meaningful insights, setting the stage for robust data exploration.
i. Importing the required Python libraries
ii. Shape of the Data frame:
Determining the shape of a DataFrame is important because it provides immediate insight into the size of the dataset, revealing the number of rows and columns.
iii. Duplicate Values:
Duplicate values in a dataset can lead to skewed or inaccurate analyses by artificially inflating the size of the data or creating misleading patterns.
iv: Info. of the Dataset:
This method includes details about the total number of entries, the number of non-null values in each column, the data types of each column (e.g., integer, float, object), and the memory usage.
v. Checking for Null Values:
Checking for null values involves identifying any missing or undefined data in each column. Understanding the presence and distribution of null values helps in making informed decisions about how to handle them, whether it’s through imputation, removal, or other methods.
vi. Head of the dataset:
vii. Tail of the dataset:
viii. Removing non-numeric characters from the columns:
In order to convert data to a numeric format, it’s often necessary to remove non-numeric characters such as currency symbols or commas. Occasionally, columns expected to contain numeric data may include non-numeric characters. The columns that underwent this cleaning process include Total Sales, Units Sold, Operating Profit, and Operating Margin.
ix. Converting Invoice Date to DateTime object:
Converting Invoice Date to a DateTime object enables precise handling of temporal data, essential for accurate date-based calculations and time-sensitive analyses.
Sales Analysis:
Total revenue: Total revenue refers to the complete amount of income generated by a business from its sales of goods or services, before any expenses are deducted. It’s a primary indicator of a company’s financial health and market performance, calculated by multiplying the price of goods or services by the total number of units sold.
Let’s start by addressing key business questions that will assist Adidas in identifying its strengths and areas for improvement.
1. Sales Performance Analysis: Which products are excelling in sales, and which are underperforming?
Key Insight:
Men’s Street Footwear emerges as the top-performing category with robust sales amounting to $208,826,244, highlighting a strong market preference. In contrast, Women’s Athletic Footwear shows a relatively lower performance, recording sales of $106,631,896.
This disparity suggests potential growth areas and indicates a need for targeted strategies to enhance the appeal of underperforming categories like Women’s Athletic Footwear.
2. Regional Market Analysis: Which stores are experiencing strong sales, and which ones are lagging?
I began by grouping the data based on two key variables: City and Retailer. This grouping allows for the examination of sales figures for specific retailers within each city. After grouping, the `.sum()` method is applied to calculate the total units sold for each retailer in each city. Finally, the grouped data is sorted for further clarity: it’s sorted in ascending order by city, ensuring that the data is presented in a systematic, city-by-city manner, and in descending order by units sold, highlighting retailers with the highest sales at the top of the list for each city.
This method offers a detailed yet straightforward way to analyze and understand the sales performance of different retailers across various cities.
Key Insight:
West Gear in Albany and Kohl’s in Albuquerque emerge as the top performers in their respective cities, with West Gear selling 47,133 units and Kohl’s selling 43,752 units.
Conversely, the lowest performers are Kohl’s in Albany and Sports Direct in Albuquerque, with sales of 10,053 and 8,881 units respectively.
3. Profit Margin Analysis: Does the Operating profit significantly impact sales?
To answer the question, we can either use detailed tables and graphs for in-depth analysis or simply assess the correlation between Operating Profit and Total Sales; a positive correlation implies a ‘yes’, while a negative one indicates a ‘no’.
Key Insight:
There’s a 0.9563 correlation between operating profit and total sales, and a 0.8923 correlation between operating profit and units sold. These high correlation values indicate that operating profit significantly impacts sales performance.
A strong operating profit is closely linked with higher total sales and a greater number of units sold, suggesting that efficient operational management plays a crucial role in driving sales success.
4. Efficiency of Sales Methods: Which sales method is more effective — in-store or online?
To tackle this question, I have utilized a library in Python called Plotly.
- First I begin by cleansing the ‘Total Sales’ data, removing currency symbols and converting the values to float for numeric analysis.
- A choropleth map is created using Plotly, with separate traces added for each unique product and sales method combination, showing total sales by U.S. state.
- The layout is enhanced with dropdown menus for product and sales method selection, enabling interactive visualization of sales data across different categories and methods.
- The final output is a dynamic, interactive visualization showing total sales across various states, differentiated by products and sales methods.
Interpretation of the table:
- Total Sales:
- In-store: $356,643,750
- Online: $247,672,882
- Outlet: $295,585,493
In-store sales have the highest total sales, followed by outlet and then online sales.
2. Operating Profit:
- In-store: $127,591,300 (approx)
- Online: $96,555,180 (approx
- Outlet: $107,988,300 (approx)
Similar to total sales, in-store sales lead in operating profit, followed by outlet and then online.
3. Operating Margin (Operating Profit divided by Total Sales):
- In-store: 0.357756 (or 35.78%)
- Online: 0.389850 (or 38.99%)
- Outlet: 0.365337 (or 36.53%)
Here, while in-store sales have the highest total sales and profit, online sales have the highest operating margin, indicating a higher profitability relative to the sales generated.
Key Insights:
In terms of Total Sales and Operating Profit: In-store sales method is the most effective, generating the highest total sales and operating profit.
In terms of Operating Margin: Online sales are the most effective, indicating that although the total sales and operating profit are lower than in-store, the profitability relative to the sales generated is higher.
When deciding which sales method is more effective, it depends on what the business prioritizes. If the focus is on maximizing total revenue and profit, then in-store sales are more effective. However, if the focus is on efficiency in terms of profit generated per dollar of sales, then online sales are more effective.
5. Price Optimization: Is there a specific price range that achieves better sales than others?
To answer this question, I analyzed the data based on different price ranges. Here’s the breakdown to my approach.
1. Defining Price Bins: I begin by creating bins for price ranges, with intervals of $10 up to $130.
2. Categorizing Prices: A new column ‘Price Range’ is added to our dataset, categorizing each product into these defined price bins based on its ‘Price per Unit’.
3. Aggregating Sales Data: I then group the data by these price ranges and calculate the sum of ‘Total Sales’ for each range.
4. Sorting for Insights: To identify which price range yields the highest sales, I sort the aggregated data in descending order of ‘Total Sales’.
5. Final Analysis: The sorted results offer a clear view of sales performance across different price ranges, enabling us to pinpoint the most and least profitable pricing tiers in our product range.
Key Insights:
The highest total sales are in the price range of $40 to $50 ($220,666,307), indicating that products priced within this range are very popular or in high demand. This could indicate a sweet spot for pricing where customers are more willing to purchase.
Very low-priced products (below $20) and higher-priced products (above $90) show significantly lower sales. This might be due to various factors like perceived value, product quality, target customer segment, or availability.
6. Product Portfolio Optimization I: Determine which products are most profitable, segmented by location.
Identifying the most profitable products for each city is crucial. By doing this, retailers can fine-tune their pricing strategies, even with slight increases in profit margins, to boost earnings. This approach benefits both the manufacturer, like Adidas, and the retailer by maximizing profitability in targeted markets.
Product popularity and profitability vary significantly by city, indicating the need for tailored marketing and sales strategies in different locations.
Key Insights:
In Albany, ‘Women’s Apparel’ and ‘Men’s Street Footwear’ are highly profitable, with profit margins around 49% and 46%, respectively.
New York excels in sales of ‘Men’s Apparel’, ‘Men’s Athletic Footwear’, and ‘Women’s Athletic Footwear’, indicating a strong market for these products.
Omaha consistently shows the lowest sales across various products, suggesting limited market penetration or demand.
7. Market Expansion Opportunities: Assess the best and worst performing stores based on their locations.
You must be wondering why is this question important to address…. Let me explain.
This approach allows businesses to make informed decisions about where to focus their growth efforts, ensuring efficient allocation of resources. By understanding how different stores perform in various locations, companies can gain a deeper insight into consumer behavior and preferences, tailor their marketing strategies accordingly, and maintain a competitive edge.
The table clearly illustrates the significant impact of profit margins on a retailer’s total sales in a city.
- The data has been aggregated to show the performance of different retail stores in various cities.
- The key metrics considered are Total Sales, Operating Profit, and Profit Margin.
- Stores are grouped by city and retailer, providing a clear view of each store’s performance in its respective location.
Key Insights:
Top Performers:
In Albany, ‘West Gear’ emerges as the top performer with a total sales of approximately $20.7 million and a profit margin of about 38.9%. This is significantly higher than ‘Kohl’s’ in the same city, which has a total sales of around $3.7 million with a profit margin of 37%.
In Albuquerque, ‘Kohl’s’ leads with a total sales of $17.1 million and a profit margin of approximately 33.9%.
2. Profit Margins Analysis:
Profit margins vary across cities and stores. While ‘West Gear’ in St. Louis has a high profit margin of 40.1%, other stores like ‘Amazon’ in Anchorage show lower profit margins (31%).
3. Sales vs. Profit Margins:
There is not always a direct correlation between high sales and high profit margins. For instance, ‘Kohl’s’ in Wichita has higher sales than ‘Foot Locker’ in the same city, yet their profit margins are fairly close (35.3% for Kohl’s vs. 34.9% for Foot Locker).
8. Time Series Analysis: Investigate whether there has been a consistent sales trend over time or any noticeable monthly trends.
The analysis of sales trends over time, based on the provided data, offers insightful observations.
Initially, the time series analysis involves grouping sales data by month and plotting these figures to visually assess trends. The resulting plot, styled for clarity and aesthetics, reveals patterns in monthly sales fluctuations. Further, a numerical analysis is conducted to determine seasonality and general sales trends. This involves calculating the month-to-month differences in sales and setting a threshold to identify significant changes. The output indicates the presence of seasonality in sales, as variations exceed the predefined threshold.
Key Insights:
This analysis suggests that while sales exhibit seasonal patterns, there’s no steady long-term growth or decline, indicating the need for a more nuanced approach to sales strategy that accounts for these periodic fluctuations.
Monthly sales data from January 2020 to January 2022 exhibit significant variability with notable peaks around April 2021 and consistent increases in December and April of each year, potentially correlating with holiday seasons and sales promotions.
9. Predictive Sales Analysis: Develop a forecast for monthly sales.
Using the ARIMA model, I’ve forecasted the next year’s monthly sales by analyzing historical trends and seasonality. This method offers more than just future projections; it’s a strategic asset for businesses. It guides efficient resource allocation, anticipates market fluctuations, and ensures adaptability in a dynamic market.
An ARIMA model with parameters (2,1,2) is fitted to the monthly sales data to forecast future trends. The model successfully predicts sales for the next 12 months, blending historical data with forecasted values. The resulting visualization, which includes both actual and forecasted sales, provides a clear depiction of expected sales performance in the upcoming year, aiding in strategic planning and resource allocation.
Key Insights:
The graph displays actual and forecasted monthly sales from January 2020 to March 2023. Actual sales peak at 7 million in April 2021 and show volatility, while forecasted sales predict a steadier pattern, mostly ranging between 4 to 6 million, without the extremes seen in actual sales. The forecasted data smoothens the peaks and valleys, indicating a stable sales environment for future planning.
10. Tableau Sales Dashboard.
Leveraging the powerful Business Intelligence capabilities of Salesforce’s Tableau, I crafted an insightful dashboard.
This dashboard provides a concise and comprehensive overview of the key insights extracted from the Adidas dataset, showcasing the valuable data-driven conclusions that can be drawn using this advanced analytical tool.
Conclusion:
This comprehensive analysis of Adidas’ sales underscores the brand’s current market position, pinpointing both its triumphs and potential growth avenues. To maintain and enhance its competitive edge in the dynamic sportswear industry, Adidas should concentrate on advancing product innovation, refining marketing strategies, bolstering its omnichannel presence, and deepening its commitment to sustainability. Armed with these data-driven insights and a steadfast focus on customer needs, Adidas is well-equipped to reinforce its status as a worldwide frontrunner, consistently offering outstanding products and experiences to its varied clientele.
Regards:
I would like to extend my gratitude to @Williams Nse Idara for inspiring this project. While I have delved into a more in-depth analysis with a focus on addressing critical questions, her work has been a valuable reference point to answer some foundational questions. You can explore her contributions through this link: Williams Nse Idara’s Project. I appreciate the foundation her work provided, and I hope my expanded analysis adds further insights to the field.
Closing Remark:
This project was a valuable learning experience for me, allowing me to harness the capabilities of both- Python and its libraries, and Tableau. Thank you for taking the time to read my article. I sincerely hope you found it as enjoyable and informative as I found working on the project.
Please feel free to share any insights or comments you have about the dashboard or the project itself.
Here are the links to my project in case if you’ve missed it- Tableau dashboard , GitHub (you can find my .ipynb file and dataset here) and my LinkedIn of course!
Thank you!