Estimating A Product’s Impact On Customer Loyalty

Estimating A Product’s Impact On Customer Loyalty
Estimating A Product’s Impact On Customer Loyalty
Estimating A Product’s Impact On Customer Loyalty
5217651242_546ba292bb_o

CATEGORY: B2C / BEAUTY
YEAR: 2017
KEYWORDS: CRM Data, Clv Model, Regression, Propensity Modeling, Counterfactual Analysis, Financial Modeling, Customer Equity, Loyalty

TYPE: B2C / BEAUTY
YEAR: 2017
KEYWORDS: Crm Data, Clv Model, Regression, Propensity Modeling, Counterfactual Analysis, Financial Modeling, Customer Equity, Loyalty

SUMMARY :

Our client was considering whether or not to keep a major product line, but was concerned about the impact of its removal on customer equity. Gradient developed a model that estimated the causal impact of this product on customer lifetime value (CLV), which allowed our client to develop a counterfactual financial model to directly inform their decision.

ASK :

Our client needed a counterfactual analysis that would give them intelligence into the long-term financial impact of removing a product line. They needed to know — at the individual customer level — what the impact of purchasing a product in this line had on customer lifetime value.

METHODOLOGY:

Gradient developed a beta-geometric beta-binomial regression model on with our client’s raw CRM data to estimate individual-level CLV. Gradient then developed a propensity model to compare individuals that had made a purchase in the target product line and those that had not, which allowed us to estimate the impact of this product line on purchasing behavior and loyalty.

RESULTS: 

The model showed that this product line was creating substantial customer equity at the individual level — customers that purchased went on to become more substantial and more loyal customers. Our client of course decided to keep the product line.

SUMMARY :

Our client was considering whether or not to keep a major product line, but was concerned about the impact of its removal on customer equity. Gradient developed a model that estimated the causal impact of this product on customer lifetime value (CLV), which allowed our client to develop a counterfactual financial model to directly inform their decision.


ASK
:

Our client needed a counterfactual analysis that would give them intelligence into the long-term financial impact of removing a product line. They needed to know — at the individual customer level — what the impact of purchasing a product in this line had on customer lifetime value.

METHODOLOGY:

Gradient developed a beta-geometric beta-binomial regression model on with our client’s raw CRM data to estimate individual-level CLV. Gradient then developed a propensity model to compare individuals that had made a purchase in the target product line and those that had not, which allowed us to estimate the impact of this product line on purchasing behavior and loyalty.


RESULTS: 

The model showed that this product line was creating substantial customer equity at the individual level — customers that purchased went on to become more substantial and more loyal customers. Our client of course decided to keep the product line.

ABOUT

We are marketers and technologists that build robust models to guide managers in their decisions. We believe in sound design, interpretability, and going out and getting the data if you don’t already have it.

ABOUT

We are marketers and technologists that build robust models to guide managers in their decisions. We believe in sound design, interpretability, and going out and getting the data if you don’t already have it.

ABOUT

We are marketers and technologists that build robust models to guide managers in their decisions. We believe in sound design, interpretability, and going out and getting the data if you don’t already have it.

ABOUT

We are marketers and technologists that build robust models to guide managers in their decisions. We believe in sound design, interpretability, and going out and getting the data if you don’t already have it.

ABOUT

We are marketers and technologists that build robust models to guide managers in their decisions. We believe in sound design, interpretability, and going out and getting the data if you don’t already have it.

Join Our Community

Join Our Community

Join Our Community

Join Our Community

COPYRIGHT GRADIENT METRICS 2018

COPYRIGHT GRADIENT METRICS 2018