Customer Lifetime Value

This regression model provides the value of each customer in a given window of time in your customer population.

How does it work?

We calculate lifetime value (LTV) using the following metrics:

• HISTORICAL_LTV: All spend prior to score date.
• SCORE: The predicted spend of a customer in the TRAINING_WINDOW.
• TOTAL_LTV: The sum of HISTORICAL_LTV and SCORE

HISTORICAL_LTV is the past spend, SCORE is the predicted future spend, and TOTAL_LTV is the combination of both. You can use all three for audience creation and segmentation, depending on your goal.

MessageGears' default value for the window of prediction (known as the PREDICTION_WINDOW) is 365 days. This means the model predicts the customer’s spend over the next 365 days. If necessary, we can assign a custom PREDICTION_WINDOW that fits a your unique sales cycle. We transform each customer’s unique data into a custom feature set, with automated training and tuning to deliver the most performant model possible.

The HISTORICAL_WINDOW determines who is eligible to be scored. For the Next Purchase model, the default HISTORICAL_WINDOW criteria is that a customer must have made a purchase, clicked an email, or triggered a web event within the last 365 days.

After training and tuning, the model is ready to score the active population.

Results

The model produces four different columns, all of which have value for marketing efforts:

• HISTORICAL_LTV: All spend prior to score date.
• SCORE: The predicted spend of a customer in the TRAINING_WINDOW.
• SCORE_SCALED: The score transformed into a ranked scale – 100 is the most valuable customers during the window of prediction, 1 is the least valuable.
• TOTAL_LTV: The sum of HISTORICAL_LTV and SCORE.

Requirements

Email engagement data (clicks, opens, sends)

• Web / Pixel engagement data
  • Data Input Schemas for Transaction/conversion data

Schema

MG Field Name	Type	Range	Description
RecipientId	VARCHAR		Unique customer identifier.
MLLTVHistoricalSpend	DECIMAL		All spend prior to score date.
MLLTVPredictedSpend	DECIMAL		The predicted spend of a customer in the TRAINING_WINDOW.
MLLTVScore	INT	1-100	The score transformed into a ranked scale – 100 is the most valuable customers during the window of prediction, 1 is the least valuable.
MLLTVTotalSpend	DECIMAL		The sum of ML-LTVHistoricalSpend (HISTORICAL_LTV) and ML-LTVPredictedSpend (SCORE).
MLLTVLow	BOOL	0 or 1	*Predefined segment where the top quintile = high
MLLTVMedium	BOOL	0 or 1	*Predefined segment where the mid three quintiles = medium
MLLTVHigh	BOOL	0 or 1	*Predefined segment where the bottom quintile = low

*High/Medium/Low LTV segments

The high/medium/low LTV segments are predefined segments. Our recommended audience definitions break out high/medium/low, where top quintile = high; mid three quintiles = medium; bottom quintile = low. These can customized this during implementation.