Hypothesis testing is sometimes used in conjunction with predictive analytics.
Lowe's Home Improvement has its own store credit cared that offers customers 90 days free financing or a 5 percent discount on purchases. Lowe's wants to know if customers who use the card have different spending habits than those that do not use the card. Lowe's is interested in answering the following questions.
Should it stop offering its own credit card?
Should it offer more or fewer perks as part of the credit card?
Should it advertise its credit card more broadly with additional customers?
It makes the following hypotheses.
Hypothesis 1: Lowe's customers using a Lowe's credit card will have a higher average sales amount than those customers not paying with a Lowe's credit card.
Hypothesis 2: Lowe's customers using a Lowe's credit card make more Lowe's visits than those customers not paying with a Lowe's credit card.
Hypothesis 3: Lowe's customers using a Lowe's credit card buy higher-margin products than those customers not paying with a Lowe's credit card.
Using statistical tests, such as a t-test of a difference in means or through regression analyses, Lowe's analysts can test for differences between the two groups of customers.