Market basket analysis working example

The following example explains the market basket analysis of a test dataset.

Raw dataset

The following is a test dataset that considers the edge cases, and is exhaustive to capture all nuances. The dataset contains 7 transactions across 3 different customers. Some transactions have only one product while some transactions have multiple products.

Raw data (Transactions)
Transaction Id	Customer Id	Product 1 name	Product 2 name	Product 1 category	Product 2 category	Product 1 label	Product 2 label	First product total quantity	Second product total quantity	Product 1 revenue	Product 2 revenue	Transaction revenue
1	111	Apple		Fruits		No discount	No discount	2		100		100
2	222	Apple	Potato	Fruits	Vegetables	15% off		3	2	150	50	350
3	111	Tshirts		Clothing		30% off		3		450		450
4	111	Apple	Rose	Fruits	Flowers	Complementary		4	2	200	400	600
5	111	Apple	Potato	Fruits	Vegetables	25% off		5	3	250	300	550
6	333	Potato	Tshirts	Vegetables	Clothing			1	1	50	150	250
7	333	Banana	Spinach	Fruits	Vegetables			2	3	100	600	700

MBA Analysis - Example 1

Report type : Transaction Level Analysis

Analysis by dimension : Product Name

Product 1 name	Product 2 name	First Product Total Quantity	Second Product Total Quantity	First Product Total Transaction Count	Second Product Total Transaction Count	Combined Total Count	First product total sales revenue	Second product total sales revenue	Combined sales revenue	First product sales revenue contribution	Second product sales revenue contribution	Support	Confidence	Lift
Apple	Potato	14	6	4	3	2	700	600	900	400	500	28.57%	50.00%	1.75
Apple	Rose	14	2	4	1	1	700	400	600	200	400	14.29%	25.00%	1.75
Potato	Tshirts	6	4	3	1	1	600	150	250	100	150	14.29%	33.33%	2.33
Banana	Spinach	2	3	1	1	1	100	600	700	100	600	14.29%	100.00%	7.00

The following is the detailed analysis:

First and second product total quantity is the aggregation of the product quantity across all transactions. Product total quantity specifies the quantity of individual products sold. 14 units of Apple are bought.
Similarly, the first and second product total transaction count is the aggregation of the transaction count. Apple appeared in 4 transactions.
First and second product total sales revenue is the aggregation of sales revenue. The revenue from Potato is $600.
Combined sales revenue is the revenue generated when both products were bought together in a transaction. Therefore, of the 7 transactions, Apple and Potato were bought together in two transactions generating a $900 revenue.
Support is the ratio of ‘combined total transaction count’ and ‘total transactions’. Here, Support for Apple and Potato is ‘2 combined transactions/7 total transactions = 28.57%’.
Confidence is the ratio of how frequently the second product was bought with the first product. Here, Confidence of Potato given that Apple was bought is ‘2 combined transactions/4 total transactions of Apple = 50%’.
A Lift score of higher than 1 indicates a strong correlation between the first and second product.

MBA Analysis - Example 2

Report type : Customer Level Analysis

Analysis by dimension : Product Name

Product category 1	Product category 2	First product category total quantity	Second product category total quantity	First product total transaction count	Second product total transaction Count	Combined total count	First product total sales revenue	Second product total sales revenue	Combined sales revenue	First product category sales revenue contribution	Second product category sales revenue contribution
Apple	Tshirts	14	4	4	2	1	700	600	550	100	450
Apple	Rose	14	2	4	1	1	700	400	500	100	400
Apple	Potato	14	6	4	3	2	700	400	900	300	600
Tshirts	Apple	4	14	2	4	2	600	700	1350	900	450
Tshirts	Rose	4	2	2	1	1	600	400	850	450	400
Rose	Apple	2	14	1	4	1	400	700	650	400	250
Rose	Potato	2	6	1	3	1	400	400	700	400	300
Potato	Banana	6	2	3	1	1	400	100	200	100	100
Potato	Spinach	6	3	3	1	1	400	600	700	100	600
Tshirts	Banana	4	2	2	1	1	600	100	250	150	100
Tshirts	Spinach	4	3	2	1	1	600	600	750	150	600
Tshirts	Potato	4	6	2	3	1	600	400	750	450	300

The following is the detailed analysis:

For customer level analysis, combinations or baskets are formed by considering subsequent transactions for a particular customer. For example, for customer id “111”, with transaction id “1”, the customer bought Apple. The next transaction of customer “111” was Tshirt with transaction id “3”. Thus, a valid basket in customer level analysis is Apple and Tshirt. Similarly, other valid combinations or baskets considering only customer id “111” can be:

(Apple and Rose) by combining transaction 1 and 4
(Apple and Potato) by combining transaction 1 and 5

Thus, Transaction Level Analysis focuses on intra-transactional combinations while Customer Level Analysis focuses on inter-transactional combinations.

MBA Analysis - Example 3

Report type : Transaction Level Analysis

Analysis by dimension : Product Category

Product Category 1	Product Category 2	First product category total quantity	Second product category total quantity	First product category total transaction count	Second product category total transaction count	Combined transaction total count	First product category total sales revenue	Second product category total sales revenue	Combined sales revenue	First product category sales revenue contribution	Second product category sales revenue contribution
Fruits	Vegetables	16	9	5	3	3	800	1200	1600	500	1100
Fruits	Flowers	16	2	5	1	1	800	400	600	200	400
Vegetables	Clothing	9	4	3	2	1	1200	600	250	100	150

The following is the detailed analysis:

The aggregation of baskets is done by Product Category.
This example explains that out of 7 transactions, Fruits and Vegetables were bought together in 3 transactions, generating a revenue of $1600, which is approximately 57% of the total revenue.

Did not find what you were looking for?

If this content did not answer your questions, try searching or contacting our support team for further assistance.

The following example explains the market basket analysis of a test dataset.

Raw dataset

Raw data (Transactions)
Transaction Id	Customer Id	Product 1 name	Product 2 name	Product 1 category	Product 2 category	Product 1 label	Product 2 label	First product total quantity	Second product total quantity	Product 1 revenue	Product 2 revenue	Transaction revenue
1	111	Apple		Fruits		No discount	No discount	2		100		100
2	222	Apple	Potato	Fruits	Vegetables	15% off		3	2	150	50	350
3	111	Tshirts		Clothing		30% off		3		450		450
4	111	Apple	Rose	Fruits	Flowers	Complementary		4	2	200	400	600
5	111	Apple	Potato	Fruits	Vegetables	25% off		5	3	250	300	550
6	333	Potato	Tshirts	Vegetables	Clothing			1	1	50	150	250
7	333	Banana	Spinach	Fruits	Vegetables			2	3	100	600	700

MBA Analysis - Example 1

Report type : Transaction Level Analysis

Analysis by dimension : Product Name

Product 1 name	Product 2 name	First Product Total Quantity	Second Product Total Quantity	First Product Total Transaction Count	Second Product Total Transaction Count	Combined Total Count	First product total sales revenue	Second product total sales revenue	Combined sales revenue	First product sales revenue contribution	Second product sales revenue contribution	Support	Confidence	Lift
Apple	Potato	14	6	4	3	2	700	600	900	400	500	28.57%	50.00%	1.75
Apple	Rose	14	2	4	1	1	700	400	600	200	400	14.29%	25.00%	1.75
Potato	Tshirts	6	4	3	1	1	600	150	250	100	150	14.29%	33.33%	2.33
Banana	Spinach	2	3	1	1	1	100	600	700	100	600	14.29%	100.00%	7.00

The following is the detailed analysis:

First and second product total quantity is the aggregation of the product quantity across all transactions. Product total quantity specifies the quantity of individual products sold. 14 units of Apple are bought.
Similarly, the first and second product total transaction count is the aggregation of the transaction count. Apple appeared in 4 transactions.
First and second product total sales revenue is the aggregation of sales revenue. The revenue from Potato is $600.
Combined sales revenue is the revenue generated when both products were bought together in a transaction. Therefore, of the 7 transactions, Apple and Potato were bought together in two transactions generating a $900 revenue.
Support is the ratio of ‘combined total transaction count’ and ‘total transactions’. Here, Support for Apple and Potato is ‘2 combined transactions/7 total transactions = 28.57%’.
Confidence is the ratio of how frequently the second product was bought with the first product. Here, Confidence of Potato given that Apple was bought is ‘2 combined transactions/4 total transactions of Apple = 50%’.
A Lift score of higher than 1 indicates a strong correlation between the first and second product.

MBA Analysis - Example 2

Report type : Customer Level Analysis

Analysis by dimension : Product Name

Product category 1	Product category 2	First product category total quantity	Second product category total quantity	First product total transaction count	Second product total transaction Count	Combined total count	First product total sales revenue	Second product total sales revenue	Combined sales revenue	First product category sales revenue contribution	Second product category sales revenue contribution
Apple	Tshirts	14	4	4	2	1	700	600	550	100	450
Apple	Rose	14	2	4	1	1	700	400	500	100	400
Apple	Potato	14	6	4	3	2	700	400	900	300	600
Tshirts	Apple	4	14	2	4	2	600	700	1350	900	450
Tshirts	Rose	4	2	2	1	1	600	400	850	450	400
Rose	Apple	2	14	1	4	1	400	700	650	400	250
Rose	Potato	2	6	1	3	1	400	400	700	400	300
Potato	Banana	6	2	3	1	1	400	100	200	100	100
Potato	Spinach	6	3	3	1	1	400	600	700	100	600
Tshirts	Banana	4	2	2	1	1	600	100	250	150	100
Tshirts	Spinach	4	3	2	1	1	600	600	750	150	600
Tshirts	Potato	4	6	2	3	1	600	400	750	450	300

The following is the detailed analysis:

(Apple and Rose) by combining transaction 1 and 4
(Apple and Potato) by combining transaction 1 and 5

Thus, Transaction Level Analysis focuses on intra-transactional combinations while Customer Level Analysis focuses on inter-transactional combinations.

MBA Analysis - Example 3

Report type : Transaction Level Analysis

Analysis by dimension : Product Category

Product Category 1	Product Category 2	First product category total quantity	Second product category total quantity	First product category total transaction count	Second product category total transaction count	Combined transaction total count	First product category total sales revenue	Second product category total sales revenue	Combined sales revenue	First product category sales revenue contribution	Second product category sales revenue contribution
Fruits	Vegetables	16	9	5	3	3	800	1200	1600	500	1100
Fruits	Flowers	16	2	5	1	1	800	400	600	200	400
Vegetables	Clothing	9	4	3	2	1	1200	600	250	100	150

The following is the detailed analysis:

The aggregation of baskets is done by Product Category.
This example explains that out of 7 transactions, Fruits and Vegetables were bought together in 3 transactions, generating a revenue of $1600, which is approximately 57% of the total revenue.

Did not find what you were looking for?

If this content did not answer your questions, try searching or contacting our support team for further assistance.

The following example explains the market basket analysis of a test dataset.

Raw dataset

Raw data (Transactions)
Transaction Id	Customer Id	Product 1 name	Product 2 name	Product 1 category	Product 2 category	Product 1 label	Product 2 label	First product total quantity	Second product total quantity	Product 1 revenue	Product 2 revenue	Transaction revenue
1	111	Apple		Fruits		No discount	No discount	2		100		100
2	222	Apple	Potato	Fruits	Vegetables	15% off		3	2	150	50	350
3	111	Tshirts		Clothing		30% off		3		450		450
4	111	Apple	Rose	Fruits	Flowers	Complementary		4	2	200	400	600
5	111	Apple	Potato	Fruits	Vegetables	25% off		5	3	250	300	550
6	333	Potato	Tshirts	Vegetables	Clothing			1	1	50	150	250
7	333	Banana	Spinach	Fruits	Vegetables			2	3	100	600	700

MBA Analysis - Example 1

Report type : Transaction Level Analysis

Analysis by dimension : Product Name

Product 1 name	Product 2 name	First Product Total Quantity	Second Product Total Quantity	First Product Total Transaction Count	Second Product Total Transaction Count	Combined Total Count	First product total sales revenue	Second product total sales revenue	Combined sales revenue	First product sales revenue contribution	Second product sales revenue contribution	Support	Confidence	Lift
Apple	Potato	14	6	4	3	2	700	600	900	400	500	28.57%	50.00%	1.75
Apple	Rose	14	2	4	1	1	700	400	600	200	400	14.29%	25.00%	1.75
Potato	Tshirts	6	4	3	1	1	600	150	250	100	150	14.29%	33.33%	2.33
Banana	Spinach	2	3	1	1	1	100	600	700	100	600	14.29%	100.00%	7.00

The following is the detailed analysis:

First and second product total quantity is the aggregation of the product quantity across all transactions. Product total quantity specifies the quantity of individual products sold. 14 units of Apple are bought.
Similarly, the first and second product total transaction count is the aggregation of the transaction count. Apple appeared in 4 transactions.
First and second product total sales revenue is the aggregation of sales revenue. The revenue from Potato is $600.
Combined sales revenue is the revenue generated when both products were bought together in a transaction. Therefore, of the 7 transactions, Apple and Potato were bought together in two transactions generating a $900 revenue.
Support is the ratio of ‘combined total transaction count’ and ‘total transactions’. Here, Support for Apple and Potato is ‘2 combined transactions/7 total transactions = 28.57%’.
Confidence is the ratio of how frequently the second product was bought with the first product. Here, Confidence of Potato given that Apple was bought is ‘2 combined transactions/4 total transactions of Apple = 50%’.
A Lift score of higher than 1 indicates a strong correlation between the first and second product.

MBA Analysis - Example 2

Report type : Customer Level Analysis

Analysis by dimension : Product Name

Product category 1	Product category 2	First product category total quantity	Second product category total quantity	First product total transaction count	Second product total transaction Count	Combined total count	First product total sales revenue	Second product total sales revenue	Combined sales revenue	First product category sales revenue contribution	Second product category sales revenue contribution
Apple	Tshirts	14	4	4	2	1	700	600	550	100	450
Apple	Rose	14	2	4	1	1	700	400	500	100	400
Apple	Potato	14	6	4	3	2	700	400	900	300	600
Tshirts	Apple	4	14	2	4	2	600	700	1350	900	450
Tshirts	Rose	4	2	2	1	1	600	400	850	450	400
Rose	Apple	2	14	1	4	1	400	700	650	400	250
Rose	Potato	2	6	1	3	1	400	400	700	400	300
Potato	Banana	6	2	3	1	1	400	100	200	100	100
Potato	Spinach	6	3	3	1	1	400	600	700	100	600
Tshirts	Banana	4	2	2	1	1	600	100	250	150	100
Tshirts	Spinach	4	3	2	1	1	600	600	750	150	600
Tshirts	Potato	4	6	2	3	1	600	400	750	450	300

The following is the detailed analysis:

(Apple and Rose) by combining transaction 1 and 4
(Apple and Potato) by combining transaction 1 and 5

Thus, Transaction Level Analysis focuses on intra-transactional combinations while Customer Level Analysis focuses on inter-transactional combinations.

MBA Analysis - Example 3

Report type : Transaction Level Analysis

Analysis by dimension : Product Category

Product Category 1	Product Category 2	First product category total quantity	Second product category total quantity	First product category total transaction count	Second product category total transaction count	Combined transaction total count	First product category total sales revenue	Second product category total sales revenue	Combined sales revenue	First product category sales revenue contribution	Second product category sales revenue contribution
Fruits	Vegetables	16	9	5	3	3	800	1200	1600	500	1100
Fruits	Flowers	16	2	5	1	1	800	400	600	200	400
Vegetables	Clothing	9	4	3	2	1	1200	600	250	100	150

The following is the detailed analysis:

The aggregation of baskets is done by Product Category.
This example explains that out of 7 transactions, Fruits and Vegetables were bought together in 3 transactions, generating a revenue of $1600, which is approximately 57% of the total revenue.

Did not find what you were looking for?

If this content did not answer your questions, try searching or contacting our support team for further assistance.

The following example explains the market basket analysis of a test dataset.

Raw dataset

Raw data (Transactions)
Transaction Id	Customer Id	Product 1 name	Product 2 name	Product 1 category	Product 2 category	Product 1 label	Product 2 label	First product total quantity	Second product total quantity	Product 1 revenue	Product 2 revenue	Transaction revenue
1	111	Apple		Fruits		No discount	No discount	2		100		100
2	222	Apple	Potato	Fruits	Vegetables	15% off		3	2	150	50	350
3	111	Tshirts		Clothing		30% off		3		450		450
4	111	Apple	Rose	Fruits	Flowers	Complementary		4	2	200	400	600
5	111	Apple	Potato	Fruits	Vegetables	25% off		5	3	250	300	550
6	333	Potato	Tshirts	Vegetables	Clothing			1	1	50	150	250
7	333	Banana	Spinach	Fruits	Vegetables			2	3	100	600	700

MBA Analysis - Example 1

Report type : Transaction Level Analysis

Analysis by dimension : Product Name

Product 1 name	Product 2 name	First Product Total Quantity	Second Product Total Quantity	First Product Total Transaction Count	Second Product Total Transaction Count	Combined Total Count	First product total sales revenue	Second product total sales revenue	Combined sales revenue	First product sales revenue contribution	Second product sales revenue contribution	Support	Confidence	Lift
Apple	Potato	14	6	4	3	2	700	600	900	400	500	28.57%	50.00%	1.75
Apple	Rose	14	2	4	1	1	700	400	600	200	400	14.29%	25.00%	1.75
Potato	Tshirts	6	4	3	1	1	600	150	250	100	150	14.29%	33.33%	2.33
Banana	Spinach	2	3	1	1	1	100	600	700	100	600	14.29%	100.00%	7.00

The following is the detailed analysis:

First and second product total quantity is the aggregation of the product quantity across all transactions. Product total quantity specifies the quantity of individual products sold. 14 units of Apple are bought.
Similarly, the first and second product total transaction count is the aggregation of the transaction count. Apple appeared in 4 transactions.
First and second product total sales revenue is the aggregation of sales revenue. The revenue from Potato is $600.
Combined sales revenue is the revenue generated when both products were bought together in a transaction. Therefore, of the 7 transactions, Apple and Potato were bought together in two transactions generating a $900 revenue.
Support is the ratio of ‘combined total transaction count’ and ‘total transactions’. Here, Support for Apple and Potato is ‘2 combined transactions/7 total transactions = 28.57%’.
Confidence is the ratio of how frequently the second product was bought with the first product. Here, Confidence of Potato given that Apple was bought is ‘2 combined transactions/4 total transactions of Apple = 50%’.
A Lift score of higher than 1 indicates a strong correlation between the first and second product.

MBA Analysis - Example 2

Report type : Customer Level Analysis

Analysis by dimension : Product Name

Product category 1	Product category 2	First product category total quantity	Second product category total quantity	First product total transaction count	Second product total transaction Count	Combined total count	First product total sales revenue	Second product total sales revenue	Combined sales revenue	First product category sales revenue contribution	Second product category sales revenue contribution
Apple	Tshirts	14	4	4	2	1	700	600	550	100	450
Apple	Rose	14	2	4	1	1	700	400	500	100	400
Apple	Potato	14	6	4	3	2	700	400	900	300	600
Tshirts	Apple	4	14	2	4	2	600	700	1350	900	450
Tshirts	Rose	4	2	2	1	1	600	400	850	450	400
Rose	Apple	2	14	1	4	1	400	700	650	400	250
Rose	Potato	2	6	1	3	1	400	400	700	400	300
Potato	Banana	6	2	3	1	1	400	100	200	100	100
Potato	Spinach	6	3	3	1	1	400	600	700	100	600
Tshirts	Banana	4	2	2	1	1	600	100	250	150	100
Tshirts	Spinach	4	3	2	1	1	600	600	750	150	600
Tshirts	Potato	4	6	2	3	1	600	400	750	450	300

The following is the detailed analysis:

(Apple and Rose) by combining transaction 1 and 4
(Apple and Potato) by combining transaction 1 and 5

Thus, Transaction Level Analysis focuses on intra-transactional combinations while Customer Level Analysis focuses on inter-transactional combinations.

MBA Analysis - Example 3

Report type : Transaction Level Analysis

Analysis by dimension : Product Category

Product Category 1	Product Category 2	First product category total quantity	Second product category total quantity	First product category total transaction count	Second product category total transaction count	Combined transaction total count	First product category total sales revenue	Second product category total sales revenue	Combined sales revenue	First product category sales revenue contribution	Second product category sales revenue contribution
Fruits	Vegetables	16	9	5	3	3	800	1200	1600	500	1100
Fruits	Flowers	16	2	5	1	1	800	400	600	200	400
Vegetables	Clothing	9	4	3	2	1	1200	600	250	100	150

The following is the detailed analysis:

The aggregation of baskets is done by Product Category.
This example explains that out of 7 transactions, Fruits and Vegetables were bought together in 3 transactions, generating a revenue of $1600, which is approximately 57% of the total revenue.

Did not find what you were looking for?

If this content did not answer your questions, try searching or contacting our support team for further assistance.