Tools for Customer Lifetime Value Estimation

Probabilistic latent customer attrition models (also known as "buy-'til-you-die models") are used to predict future purchase behavior of customers. This package includes fast and accurate implementations of various probabilistic latent customer attrition models for non-contractual settings (e.g., retail business) with and without time-invariant and time-varying covariates. Currently, the package includes the Pareto/NBD model (Pareto/Negative-Binomial-Distribution), the BG/NBD mode (Beta-Gamma/Negative-Binomial-Distribution) and the GGom/NBD (Gamma-Gompertz/Negative-Binomial-Distribution) for the purchase and the attrition processes as well as the Gamma/Gamma model for the spending process. For reference to the Pareto/NBD model, see Schmittlein DC, Morrison DG, Colombo R (1987) , for the BG/NBD model, see Fader PS, Hardie BG, Lee K (2005) and for the GGom/NBD model see Bemmaor AC, Glady N (2012) . For reference to the Gamma/Gamma model, see Fader PS, Hardie BG, Lee K (2005) .


Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.


0.6.0 by Patrick Bachmann, 2 months ago

Report a bug at

Browse source code at

Authors: Patrick Bachmann [cre, aut] , Jeffrey Naef [aut] , Patrik Schilter [aut] , Markus Meierer [aut] , Niels Kuebler [aut]

Documentation:   PDF Manual  

GPL-3 license

Imports data.table, foreach, ggplot2, lubridate, Matrix, MASS, optimx, stats, utils

Depends on methods

Suggests BTYD, covr, doFuture, doParallel, future, knitr, rmarkdown, testthat

Linking to Rcpp, RcppArmadillo, RcppGSL

System requirements: C++11

See at CRAN