If the Model Fits, Use It: Methods and Benchmarks for Evaluating NBD-Dirichlet Goodness-of-Fit

Driesener, Carl; Banelis, Melissa; Rungie, Cam

doi:10.1016/j.ausmj.2017.10.003

Cited by 16 publications

(36 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One questionable aspect of previous Dirichlet model deviations is the seemingly arbitrary 10% deviation benchmark (Anesbury, Nguyen, & Bogomolova, 2018;Kahn et al, 1988;Scriven et al, 2017). Our research, while using this approach, provides additional support for the work of Driesener et al (2017) who proposed using eight goodness-of-fit measures to provide a more robust determination of Dirichlet modeling.…”

Section: Theoretical Implicationsmentioning

confidence: 52%

“…The aggregation is a very important feature of the Dirichlet (Goodhardt et al, 1984) that allows the analysis of behavior and ensures correct comparison of results. As demonstrated in this study, the feature goes beyond the aggregation of small brands as "other" (Driesener et al, 2017;Ehrenberg et al, 2004;Graham et al, 2017;Scriven et al, 2017).…”

Section: Data Descriptionmentioning

confidence: 86%

“…), so we use a secondary approach. Driesener et al (2017) propose using eight goodness-of-fit tests to determine if Dirichlet accurately models the brand buying behavior. The tests compare the observed and Dirichlet theoretical values for both penetration and average purchase frequency in four ways.…”

Section: Dj-a Behavioral Loyalty Patternmentioning

confidence: 99%

“…Given the 10% deviation benchmark limitation (Kahn et al, 1988), the study uses the eight goodness-of-fit tests (Driesener et al, 2017) to determine if the fresh category buying is in a Dirichlet manner. The tests involve the comparison between the observed and Dirichlet theoretical values for both penetration and average purchase frequency in four ways-comparisons of the averages, correlations, MAD, and MAPE.…”

Section: Data Descriptionmentioning

confidence: 99%

“…Employing Driesener et al (2017) tests of fit, using the data of Figure 1-tomatoes disaggregate analysis for Retailer 2 was 8/8 (minimum 5/8). Table 3 shows 76% (46 of 60) analyses, and the categories have five or more tests passed.…”

Section: Djmentioning

confidence: 99%

See 4 more Smart Citations

Analyzing proprietary, private label, and non-brands in fresh produce purchases

Anesbury

Jürkenbeck

Bogomolov

et al. 2020

International Journal of Market Research

View full text Add to dashboard Cite

When purchasing packaged products within a supermarket, consumers choose between proprietary or private label brands. However, when purchasing fresh fruits and vegetables, non-branded produce is the dominant option—with proprietary and private label brands only recently becoming available. Previous fast-moving consumer goods (FMCG) research finds that proprietary and private label brands affect consumer loyalty—however, no research exists for fresh categories. This research is the first to determine the effect of emerging brands in fresh categories on consumer buying behavior. Our research examines consumers’ loyalty toward proprietary, private label, or non-branded fresh fruits and vegetables and the level of customer sharing between these options, using analytical approaches applicable to FMCG categories. The panel data contains nearly 46,000 households making over 8 million purchases in the United States during 2015. Results show that proprietary, private label, and now non-branded fresh produce have expected loyalty levels, for their size, and consumers share their purchases across the three options (i.e., consumers are not loyal to just one option). The study analyzes and interprets purchase data in fresh categories offering marketing academics and practitioners actionable advice for working with fresh produce purchase data.

show abstract

Section: Theoretical Implicationsmentioning

confidence: 52%

Section: Data Descriptionmentioning

confidence: 86%

Section: Dj-a Behavioral Loyalty Patternmentioning

confidence: 99%

Section: Data Descriptionmentioning

confidence: 99%

Section: Djmentioning

confidence: 99%

See 3 more Smart Citations

Analyzing proprietary, private label, and non-brands in fresh produce purchases

Anesbury

Jürkenbeck

Bogomolov

et al. 2020

International Journal of Market Research

View full text Add to dashboard Cite

show abstract

The Dirichlet model in marketing

Driesener

Rungie

2021

J of Consumer Behaviour

Self Cite

View full text Add to dashboard Cite

We present a review of Gerald Goodhardt's most famous contribution to marketing science-the NBD-Dirichlet model. This provides a powerful illustration of the complex pathway and useful associated discoveries that over 25 years lead to the specification and application of a key marketing model. We identify the process that started with the negative binomial distribution (NBD) applied to purchase incidence, examined alternatives such as the logarithmic series distribution (LSD) and the beta binomial distribution (BBD), and along the way developed conditional trend analysis (CTA). We then explore the development of brand choice beginning with duplication of purchase as a method for understanding the underlying patterns, before moving into various approaches to modelling choice including a multivariate NBD, and eventually a multivariate BBD-the Dirichlet multinomial distribution (DMD). We discuss key events in modelling consumer behaviour and outline the model's implications for how marketers should think about consumers, or more specifically the Dirichlet consumer. Finally, we provide a survey of the model's applications uncovering a rich recent history and bright future for Goodhardt's legacy.

show abstract

How websites compete in the Middle East: The example of Iran

Naami

Anesbury

Stocchi

et al. 2021

J of Consumer Behaviour

View full text Add to dashboard Cite

Based on the analysis of two sets of data (a cross‐sectional online survey of five product categories with an average sample size of 525 and a longitudinal telecommunications panel of more than two million respondents), this study detects a positive relationship between the market size (purchase penetration) of Iranian e‐brands (or websites) and the percentage of customers shared with other e‐brands. This finding is consistent with the well‐established Duplication of Purchase Law; it also holds over time and across different markets (e.g., repertoire vs. subscription). Hence, this study makes a twofold contribution to marketing knowledge. First, it expands the collection of empirical evidence concerning the Duplication of Purchase, which thus far is primarily within offline contexts and Western countries. Second, it addresses issues inherent to research on e‐loyalty, such as the over emphasis on evaluating loyalty for one e‐brand at a time via complex attitudinal measures. Accordingly, this study advances consumer buying behaviour research by clarifying that, similar to offline domains and other geographical areas, e‐loyalty in this buoyant Middle Eastern market divides across a small number of e‐brands. It is also best appraised through behavioural loyalty and by comparing multiple e‐brands competing within the same market. These outcomes translate into a series of practical guidelines for the strategic management of e‐brands, improving the practical understanding how e‐brands compete and grow.

show abstract

If the Model Fits, Use It: Methods and Benchmarks for Evaluating NBD-Dirichlet Goodness-of-Fit

Cited by 16 publications

References 32 publications

Analyzing proprietary, private label, and non-brands in fresh produce purchases

Analyzing proprietary, private label, and non-brands in fresh produce purchases

The Dirichlet model in marketing

How websites compete in the Middle East: The example of Iran

Contact Info

Product

Resources

About