Paper: arXiv 2506.22711

Authors: João B. G. de Brito, Rodrigo Heldt, Cleo S. Silveira, Matthias Bogaert, Guilherme B. Bucco, Fernando B. Luce, João L. Becker, Filipe J. Zabala, Michel J. Anzanello

Abstract

Financial institutions increasingly adopt customer-centric strategies to enhance profitability and build long-term relationships. While Customer Lifetime Value (CLV) is a core metric, its calculations often rely solely on single-entity data, missing insights from customer activities across multiple firms. This study introduces the Potential Customer Lifetime Value (PCLV) framework, leveraging Open Banking (OB) data to estimate customer value comprehensively. We predict retention probability and estimate Potential Contribution Margins (PCM) from competitor data, enabling PCLV calculation. Results show that OB data can be used to estimate PCLV per competitor, indicating a potential upside of 21.06% over the Actual CLV. PCLV offers a strategic tool for managers to strengthen competitiveness by leveraging OB data and boost profitability by driving marketing efforts at the individual customer level to increase the Actual CLV.

Complexity vs Empirical Score

  • Math Complexity: 3.0/10
  • Empirical Rigor: 7.5/10
  • Quadrant: Street Traders — empirical and applied, lower math complexity

Why this score: The paper applies standard machine learning models (XGBoost) and conventional CLV formulas with minimal novel mathematical derivations, focusing on data integration and model performance metrics. Empirical rigor is high due to detailed methodology on model training, cross-validation, and use of real-world Open Banking data with specific performance metrics (RMSE, MAE, PR-AUC).

Research Flowchart

  flowchart TD
  A["Research Goal: <br>Leverage Open Banking data <br>to estimate Potential CLV"] --> B["Data Collection<br>Open Banking Data<br>Transaction History"]
  B --> C["Methodology: <br>PCLV Framework<br>Retention Prediction & PCM Estimation"]
  C --> D["Compute Potential Contribution Margins<br>from Competitor Data"]
  C --> E["Predict Retention Probability<br>using OB Insights"]
  D & E --> F["Calculate Potential CLV<br>(PCLV)"]
  F --> G["Key Finding: <br>PCLV shows 21.06% upside <br>over Actual CLV"]
  G --> H["Outcome: <br>Strategic tool for targeted marketing<br>& boosting profitability"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes the Potential Customer Lifetime Value (PCLV) framework that uses Open Banking data to estimate customer value beyond a single institution.

Source sentence
This study introduces the Potential Customer Lifetime Value (PCLV) framework, leveraging Open Banking (OB) data to estimate customer value comprehensively.

Assumes 4

  • A 12‑month prediction horizon is used for PCM and PCLV calculations. model
    Source sentence
    assuming a prediction horizon of 12 months, as presented in Eq. 5.
  • An annual discount rate of 0.1 is applied in PCLV valuation. model
    Source sentence
    A 0.1 annual discount rate was applied, with monetary values converted to USD at R$4.8521/$1 on Dec 31, 2023.
  • PCM for the next 12 months is projected by multiplying the predicted monthly PCM by 12, assuming the last training month’s PCM persists. model
    Source sentence
    Multiplying the PCM by 12 projects the estimated monthly PCM over the next 12 months, assuming the PCM from the last training month, reasonably predicts the upcoming year's value.
  • Only customers who share OB data can have PCM and PCLV predicted. data
    Source sentence
    only 10,507 customers (0.32%) shared OB data, enabling PCM and PCLV predictions exclusively for these customers.

Data

  • Brazilian bank credit‑related PSC dataset · major Brazilian bank alternative proprietary
    Jan 2021–Dec 2023 · 3,335,034 customers, 261 million transactions
    Source sentence
    A 36-month dataset (Jan 2021–Dec 2023) covering 261 million transactions for 3,335,034 customers was processed for churn prediction
  • Open Banking competitor credit data alternative unknown
    Dec 2023 · 10,507 customers who shared OB data
    Source sentence
    However, only 10,507 customers (0.32%) shared OB data, enabling PCM and PCLV predictions exclusively for these customers.

Costs

Not applicable The paper reports no trading or execution result.

Finds 3

  • The churn prediction model achieved a PR‑AUC of 0.96, sensitivity 0.88, specificity and accuracy of 0.92. predictive-accuracy
    Source sentence
    Using 10-fold cross-validation, it achieved a PR-AUC of 0.96, sensitivity of 0.88, and specificity and accuracy of 0.92.
  • The XGBoost regression model for PCM prediction yielded a mean RMSE of $74 and MAE of $11.13. predictive-accuracy
    Source sentence
    resulting in a mean RMSE of $74 and MAE of $11.13, demonstrating its predictive accuracy.
  • Potential Customer Lifetime Value (PCLV) provides an estimated upside of 21.06% over Actual CLV. performance
    Source sentence
    Results show that OB data can be used to estimate PCLV per competitor, indicating a potential upside of 21.06% over the Actual CLV.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 20 related papers). Coverage is partial, so this is neither replication nor refutation.

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.

Cited References (21)

  1. Benchmarking Machine Learning Techniques for Bankruptcy Prediction under Benign and Adversarial Behaviors (2024) — link
  2. A framework to improve churn prediction performance in retail banking (2024) — link
  3. Open Banking: Definition and Description (2021) — link
  4. Forecasting gold price with the XGBoost algorithm and SHAP interaction values (2021) — link
  5. Emerging Market Retail: Transitioning from a Product-Centric to a Customer-Centric Approach (2021) — link
  6. The evolution and future of data-driven finance in the EU (2020) — link
  7. Predicting customer value per product: From RFM to RFM/P (2019) — link
  8. Creating Enduring Customer Value (2016) — link
and 13 more
  1. XGBoost: A Scalable Tree Boosting System (2016) — link
  2. Creating Value with Big Data Analytics: Making Smarter Marketing Decisions (2016) — link
  3. Marketing Metrics: The Manager’s Guide to Measuring Marketing Performance (2015)
  4. Practical Bayesian Optimization of Machine Learning Algorithms (2012) — link
  5. Expanding the Role of Marketing: From Customer Equity to Market Capitalization (2009) — link
  6. Counting Your Customers the Easy Way: An Alternative to the Pareto/NBD Model (2005) — link
  7. Managing Customers as Investments: The Strategic Value of Customers in the Long Run (2005)
  8. Open Banking: Opportunities and Risks (2023) — link
  9. Dynamics of customer segments: A predictor of customer lifetime value (2021) — link
  10. The Customer Centricity Playbook: Implement a Winning Strategy Driven by Customer Lifetime Value (2018)
  11. EUR-Lex 2015/2366 (2015)
  12. Strategic brand management, 2nd edition (2011)
  13. The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail (2005) — link

0 of 21 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.