Overview
The Market Basket feature of Resto iQ Analytics provides the user with the power to understand how the product items are performing relative to each other, which goes well with another item, and which are best to be bundled instead.
Example
If the customer orders "Product A", what is the likelihood and probability that "Product B" is going to be ordered as well? What is the probability differences of B with "Product C", etc. This information can help an organization identify how to group certain items together to either increase revenue, reduce cost, and/or increase market share.
Definitions
 Support
 Definition: How often an item or combination of items are ordered.
 Example: If the coffee is bought 20x out of 100 transactions, then the Support is 0.2 (or 20% in percentage format).  Lift
 Definition: How much more likely items are bought together than separately.
 Example: If buying Coffee and Croissant together happens twice as often as we would expect by chance, then the Lift is 2.0 (or 200%).  Confidence
 Definition: The likelihood of buying an item given that another item is bought.
 Example: If out of 200 transactions of Carrot Cake, a cup of Brewed Coffee is ordered 50x, then the Confidence is 0.25 (or 25%).  Conviction
 Definition: How strongly the absence of one item predicts the absence of another.
 Example: If Hot Tea is 1.5 times more likely to be bought without a Brewed Coffee than if they were independent, the Conviction is 1.5 (or 150%)
Antecedent v Consequent

Antecedent
Definition: The item or items that come first in a rule.
Example: In the rule Coffee → Croissant, Coffee is the antecedent.

Consequent
Definition: The item or items that come second in a rule.
Example: In the rule Coffee → Croissant, Croissant is the consequent.
User Interface
Identical with other parts of Resto iQ, Filters are placed on top of the reports which are found below.
The reports in the Market Basket module are a series of radar charts using the data found at the very bottom, similar to this:
Example of possible interpretations:
Support:
The radar graph for support indicates that the “Croissant” and “Cold Brew Coffee” are more frequently bought together compared to “Hot Tea,” which suggests a stronger association in sales between these two drinks.
Lift:
The lift graph shows that the purchase of “Croissant” and “Cold Brew Coffee” significantly increases the likelihood of one another being bought together compared to random chance, highlighting a potentially profitable bundling opportunity for these items.
Confidence:
The radar graph for confidence shows that when “Croissant” is bought, there is a high likelihood that “Cold Brew Coffee” is also purchased, indicating a strong association between these items compared to “Hot Tea.”
Conviction:
The conviction graph illustrates that transactions including “Croissant” significantly decrease the chance of “Cold Brew Coffee” being bought independently, suggesting a robust predictive relationship between these items.
Table
Lift:
Lift measures the increase in the likelihood of purchasing the consequent item when the antecedent is purchased. For instance, buying “Croissant” makes it 3.12 times more likely that “Cold Brew Coffee” will also be bought.
Confidence:
Confidence indicates the percentage of transactions with the antecedent that also include the consequent. For example, 13.34% of purchases that include “Croissant” also include “Cold Brew Coffee.”
Support:
Support shows how frequently the items appear together in transactions. For example, “Croissant” and “Cold Brew Coffee” appear together in 1.05% of all transactions.
Conviction:
Conviction measures how strongly the absence of the consequent would imply the absence of the antecedent. For instance, if “Croissant” is not bought, it is 13.1 times less likely that “Cold Brew Coffee” is not bought, indicating a strong association between the two items.