DATA
MODELLING &
Group
Aashish Changra 61KB
Aditya Kumar 63KB
VISUALISATION Parul Tebriwal 95KB
Shaiza Khanam 105KB
Shishir Kumar 108KB
AMAZON RETAIL DATASET
The Amazon retail dataset contains rich information about product sales, inventory,
customer demographics, store locations, and dates.
KEY FOCUS AREAS
Sales trends over time Customer demographics
Product category performance Store-level performance
Table Name Attributes
SalesID, DateKey,ProductKey, StoreKey, Key Relationships Between Tables
FactSales: Tracks individual sales transactions CustomerKey, Quantity Sold, Sales
Amount, Discount FactSales:
InventoryID, DateKey, ProductKey,
DateKey → DimDate
FactInventory: Tracks daily inventory levels
StoreKey, UnitOnHand, UnitsReceived ProductKey → DimProduct
StoreKey → DimStore
DimProduct: Contains information about the Product Name, Category (e.g., Fashion,
products being sold Home & Kitchen), Brand, Price.
CustomerKey → DimCustomer
DimStore: Stores and their geographical
Store Name, City, State FactInventory:
locations
DateKey → DimDate
DimCustomer: Provides customer ProductKey → DimProduct
Thynk Unlimited
Gender, Age, Location
demographics
StoreKey → DimStore
DimDate: Provides date-related details Year, Month, Quarter, Day of Week
AMAZON RETAIL DATASET
The Amazon retail dataset contains rich information about product sales, inventory,
customer demographics, store locations, and dates.
KEY INSIGHTS ISSUES IDENTIFIED
Inventory Stockouts:
Top-Performing Stores: Stores like Amazon NYC Issue: Fast-moving products like electronics often
and Amazon LA generated the highest revenue out of stock.
(20% above average). Impact: Missed sales opportunities, lower customer
Inventory Issues: Stockouts in rural stores (e.g., satisfaction.
Amazon San Antonio) affected sales. Recommendation: Improve inventory forecasting
High Discounts, Low Sales: Stores like Amazon and use Power BI alerts for low stock.
Phoenix offered large discounts but saw low sales
growth. Misaligned Discount Strategies:
Customer Segmentation: Younger demographics Issue: High discounts did not drive sales in certain
bought more tech products in Amazon LA. stores.
Impact: Lost revenue despite heavy promotions.
Recommendation: Adjust discounts based on
product demand in each region.
Model View
Dashboard
Sales Trend Overtime (Top-left chart)
Insight: The sales amount (blue line) shows
significant fluctuations throughout the year. There
are some distinct spikes in sales, suggesting
promotional periods or events that drive high sales
volume.
Recommendation
Discount (purple line) is fairly steady but shows
Investigate the dates with the highest sales spikes periods where discounting slightly increased,
and compare them with marketing campaigns or potentially aligning with the sales spikes. These could
holidays. If the spikes align, replicating those indicate seasonal sales or marketing-driven events
strategies could be beneficial for future sales (e.g., Black Friday, Prime Day).
planning.
Sales by Product Category (Bottom-left bar chart)
Insights
Category SalesAmount
Fashion 278.72
Home & Kitchen 266.23
Electronics 261.73
Books 258.8
Despite being a close competitor, Books are lagging slightly compared to other categories.
Recommendation
Fashion clearly dominates, but the gap is small between other categories. Consider
cross-promotions between Fashion and other underperforming categories to
encourage purchases across categories.
Sales by Customer Demographics (Middle chart)
Insights
The gender distribution is almost perfectly balanced,
with 50% female customers and 49% male customers.
This suggests that the product offering appeals equally
to both genders, which is great for maintaining broad
market appeal.
Recommendation
Since the gender ratio is balanced, you can
maintain neutral marketing strategies but
might want to dig deeper into specific
product preferences by gender to target
promotions better.
Sales by Brands & Categories (Top-right bar chart)
Insights
Electronics (AmazonBasics and Fire
TV) are performing very well in
terms of sales.
Fashion has steady sales across
brands, while Books and Home &
Kitchen categories have scattered
brand performance.
Kindle, Echo, and Ring (electronics)
also show strong performance,
indicating an affinity for Amazon-
branded electronics.
Recommendation
Continue investing in AmazonBasics and Fire TV marketing campaigns, as they are
performing better than other brands. Books and Home & Kitchen categories can
benefit from focused brand development to increase their share.
Store Performance (Inventory) (Top-right box)
Insights
The goal was 5,000 units of inventory, but only
2.33 units are left, representing a -99.95%
deviation.
This suggests either a stock-out situation or
insufficient supply chain replenishment.
Recommendation
Conduct an inventory health check immediately.
Investigate if this drop is seasonal or indicates poor
forecasting and stock management. Preventing
stockouts should be a high priority as it impacts sales
and customer satisfaction.
Average of Discount and Discount Target by Date (Top-right box)
Insights
The average discount (34.33) has significantly exceeded
the target of 20, with a +71.67% increase.
Offering higher discounts might have contributed to
short-term sales but could impact profitability.
Recommendation
Re-evaluate discounting strategies to strike a balance
between driving sales and maintaining profitability.
Consider personalized discounts or loyalty-based
rewards to retain margins.
Inventory Levels by Store (Bottom-left table)
Insights Recommendation
Amazon NYC has the highest Focus inventory replenishment efforts on
inventory on hand (885 units) and Amazon Houston, as it may need more
units received (409 units), indicating attention to avoid stock-outs. Consider
it's the largest or most active store. reallocating stock from stores like
Amazon Houston has the lowest Amazon NYC, which may be
inventory levels, with both units on overstocked compared to others.
hand and units received falling behind
other stores.
Average of Sales Amount and Sales Target by Date (Bottom-right box)
Insights
The actual sales amount (324.67) is
significantly below the target of 1.07K,
showing a -69.7% deviation.
Recommendation
Analyze whether the sales targets were too
ambitious or if there were specific reasons (e.g.,
stockouts, low traffic) that led to underperformance.
Adjust marketing strategies, discounts, and
promotions to bridge this gap.
