Extending the BG/NBD: A simple model of purchases and complaints

R.D. van Oest, G.A.H. Knox

Research output: Contribution to journalArticleScientificpeer-review

Abstract

Extant customer-base models like the beta geometric/negative binomial distribution (BG/NBD) predict future purchasing based on customers' observed purchase history. We extend the BG/NBD by adding an important non-transactional element that also drives future purchases: complaint history. Our model retains several desirable properties of the BG/NBD: it can be implemented in readily available software, and estimation requires only customer-specific statistics, rather than detailed transaction-sequence data. The likelihood function is closed-form, and managerially relevant metrics are obtained by drawing from beta and gamma densities and transforming these draws to a sample average. Based on more than two years of individual-level data from a major U.S. internet and catalog retailer, our model with complaints outperforms both the original BG/NBD and a modified version. Even though complaints are rare and non-transactional events, they lead to different substantive insights about customer purchasing and drop-out: customers purchase faster but also drop out much faster. Furthermore, there is more heterogeneity in drop-out rates following a purchase than a complaint.
Original languageEnglish
Pages (from-to)30-37
JournalInternational Journal of Research in Marketing
Volume28
Issue number2
Publication statusPublished - 2011

Fingerprint

Complaints
Negative binomial
Purchase
Drop out
Purchasing
Statistics
World Wide Web
Software
Retailers

Cite this

@article{effef6d6fa5f4abd948310eaf38db85a,
title = "Extending the BG/NBD: A simple model of purchases and complaints",
abstract = "Extant customer-base models like the beta geometric/negative binomial distribution (BG/NBD) predict future purchasing based on customers' observed purchase history. We extend the BG/NBD by adding an important non-transactional element that also drives future purchases: complaint history. Our model retains several desirable properties of the BG/NBD: it can be implemented in readily available software, and estimation requires only customer-specific statistics, rather than detailed transaction-sequence data. The likelihood function is closed-form, and managerially relevant metrics are obtained by drawing from beta and gamma densities and transforming these draws to a sample average. Based on more than two years of individual-level data from a major U.S. internet and catalog retailer, our model with complaints outperforms both the original BG/NBD and a modified version. Even though complaints are rare and non-transactional events, they lead to different substantive insights about customer purchasing and drop-out: customers purchase faster but also drop out much faster. Furthermore, there is more heterogeneity in drop-out rates following a purchase than a complaint.",
author = "{van Oest}, R.D. and G.A.H. Knox",
year = "2011",
language = "English",
volume = "28",
pages = "30--37",
journal = "International Journal of Research in Marketing",
issn = "0167-8116",
publisher = "Elsevier Science BV",
number = "2",

}

Extending the BG/NBD : A simple model of purchases and complaints. / van Oest, R.D.; Knox, G.A.H.

In: International Journal of Research in Marketing, Vol. 28, No. 2, 2011, p. 30-37.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Extending the BG/NBD

T2 - A simple model of purchases and complaints

AU - van Oest, R.D.

AU - Knox, G.A.H.

PY - 2011

Y1 - 2011

N2 - Extant customer-base models like the beta geometric/negative binomial distribution (BG/NBD) predict future purchasing based on customers' observed purchase history. We extend the BG/NBD by adding an important non-transactional element that also drives future purchases: complaint history. Our model retains several desirable properties of the BG/NBD: it can be implemented in readily available software, and estimation requires only customer-specific statistics, rather than detailed transaction-sequence data. The likelihood function is closed-form, and managerially relevant metrics are obtained by drawing from beta and gamma densities and transforming these draws to a sample average. Based on more than two years of individual-level data from a major U.S. internet and catalog retailer, our model with complaints outperforms both the original BG/NBD and a modified version. Even though complaints are rare and non-transactional events, they lead to different substantive insights about customer purchasing and drop-out: customers purchase faster but also drop out much faster. Furthermore, there is more heterogeneity in drop-out rates following a purchase than a complaint.

AB - Extant customer-base models like the beta geometric/negative binomial distribution (BG/NBD) predict future purchasing based on customers' observed purchase history. We extend the BG/NBD by adding an important non-transactional element that also drives future purchases: complaint history. Our model retains several desirable properties of the BG/NBD: it can be implemented in readily available software, and estimation requires only customer-specific statistics, rather than detailed transaction-sequence data. The likelihood function is closed-form, and managerially relevant metrics are obtained by drawing from beta and gamma densities and transforming these draws to a sample average. Based on more than two years of individual-level data from a major U.S. internet and catalog retailer, our model with complaints outperforms both the original BG/NBD and a modified version. Even though complaints are rare and non-transactional events, they lead to different substantive insights about customer purchasing and drop-out: customers purchase faster but also drop out much faster. Furthermore, there is more heterogeneity in drop-out rates following a purchase than a complaint.

M3 - Article

VL - 28

SP - 30

EP - 37

JO - International Journal of Research in Marketing

JF - International Journal of Research in Marketing

SN - 0167-8116

IS - 2

ER -