Customer Lifetime Yield Impact (CLYI)

GMI-MKT-PPO-018Marketing: Pricing and Promotional OptimizationIntegration-Required

Definition

The Customer Lifetime Yield Index measures whether AI-optimized pricing strategies are sustaining or eroding long-term customer value by comparing the lifetime value of customers acquired or retained under peak dynamic pricing against customers acquired or retained under stable pricing conditions.

The Measurement Gap

Measurement Gap and Decision

We propose this metric because AI pricing engines optimize for transaction-level margin, which creates a systematic blind spot: the algorithm captures maximum short-term value from each transaction without measuring whether that extraction is training customers to distrust the brand, wait for lower prices, or defect to competitors with more predictable pricing. CLYI closes that blind spot by tracking whether peak-priced customers generate comparable lifetime value to stable-priced customers over cohort time horizons.

Legacy customer lifetime value (CLV) models calculate an aggregate number: the expected total revenue from a customer over the relationship. That aggregate hides the pricing strategy's contribution. Two customer cohorts might have identical CLV projections today, but the cohort acquired through aggressive dynamic pricing may show accelerating churn at month 18 while the cohort acquired through stable pricing shows steady retention. Legacy CLV does not decompose the pricing signal from the product signal, the service signal, and the acquisition channel signal.

The decision CLYI enables is pricing strategy constraint. When CLYI drops below the proposed 0.9 threshold (meaning peak-priced customers generate less than 90% of the lifetime value of stable-priced customers) the pricing team has evidence that short-term margin optimization is destroying long-term value. The appropriate response is not to abandon dynamic pricing, but to constrain it: set ceiling parameters, reduce price volatility in segments where CLYI is declining, or adjust the algorithm's objective function to include a lifetime value penalty. Without CLYI, pricing teams celebrate transaction-level margin gains while customer retention teams fight an undiagnosed churn problem that originates in the pricing engine.

Formula

Primary Formula:

CLYI = Average LTV of Peak-Priced Cohort / Average LTV of Stable-Priced Cohort

Where:

  • Peak-Priced Cohort = customers whose acquisition or key retention transaction occurred when the AI pricing engine was operating above the 75th percentile of its dynamic range for that product or segment
  • Stable-Priced Cohort = customers whose equivalent transaction occurred when the AI pricing engine was operating within the 25th to 75th percentile of its dynamic range
  • LTV = customer lifetime value calculated at a consistent time horizon (we propose 24-month LTV for comparability across cohorts of different ages)

Critical caveat: cohort matching is required.

Customers who buy at peak prices are not a random sample. They may be more urgent, less price-sensitive, more likely to be one-time purchasers, or acquired through different channels than customers who buy at stable prices. Comparing raw LTV across these groups conflates the pricing effect with pre-existing customer differences. Before computing CLYI, match peak-priced and stable-priced cohorts on acquisition channel, customer segment, purchase history depth, and product category. The matched comparison isolates the pricing effect from the customer composition effect. Without matching, a CLYI of 0.82 could mean "peak pricing costs us 18% of lifetime value" or "peak pricing attracts customers who were always going to be lower-LTV." The distinction determines whether the remedy is a pricing change or an acquisition strategy change.

Computation Notes:

Cohort definition is the critical design decision. The peak-priced cohort should represent customers who experienced the most aggressive algorithmic pricing, not merely customers who paid the highest price. A customer who paid a high price for a premium product at list is not in the peak-priced cohort. A customer who paid a normal product's price at the top of its dynamic range because the algorithm detected urgency signals -- that is the peak-priced cohort.

LTV should include revenue, margin, and behavioral signals: purchase frequency, average order value trajectory, return rate, and engagement metrics. A customer who continues buying but at declining frequency and value is exhibiting attrition that revenue-only LTV may not capture until the relationship is effectively over.

Proposed formula, pending validation pilots.

At a Glance

Unit
Ratio
Trend Direction
Higher; cohort parity
Reporting Frequency
Annually
Category
Pricing Governance and Human-AI Calibration
← Cross-System Price Coherence Score (CSPCS)Pricing Override Value Index (OVI-MKT) →

Assess Your Organization

The AI-GPS Assessment evaluates your AI governance maturity and recommends KPIs like this one for your implementation roadmap.

Take the Free AI-GPS Assessment