Consider or Choose? The Role and Power of Consideration Sets
Consideration sets play a crucial role in discrete choice modeling, where customers often form consideration sets in the first stage and then use a second-stage choice mechanism to select the product with the highest utility. While many recent studies aim to improve choice models by incorporating more sophisticated second-stage choice mechanisms, this paper takes a step back and goes into the opposite extreme. We simplify the second-stage choice mechanism to its most basic form and instead focus on modeling customer choice by emphasizing the role and power of the first-stage consideration set formation. To this end, we study a model that is parameterized solely by a distribution over consideration sets with a bounded rationality interpretation. Intriguingly, we show that this model is characterized by the axiom of symmetric demand cannibalization, enabling complete statistical identification. The latter finding highlights the critical role of consideration sets in the identifiability of two-stage choice models. We also examine the model's implications for assortment planning, proving that the optimal assortment is revenue-ordered within each partition block created by consideration sets. Despite this compelling structure, we establish that the assortment problem under this model is NP-hard even to approximate, highlighting how consideration sets contribute to nontractability, even under the simplest uniform second-stage choice mechanism. Finally, using real-world data, we show that the model achieves prediction performance comparable to other advanced choice models. Given the simplicity of the model's second-stage phase, this result showcases the enormous power of first-stage consideration set formation in capturing customers' decision-making processes.
Code (0)
등록된 구현이 없습니다.
Tasks
Assortment OptimizationDecision MakingMarketingSimilar Papers 제목 키워드 기반
When the Universe is Too Big: Bounding Consideration Probabilities for Plackett-Luce Rankings
The widely used Plackett-Luce ranking model assumes that individuals rank items by making repeated choices from a universe of items. But in many cases the universe is too big for people to plausibly consider all options.…
Towards rational glyco-engineering in CHO: from data to predictive models
Metabolic modeling strives to develop modeling approaches that are robust and highly predictive. To achieve this, various modeling designs, including hybrid models, and parameter estimation methods that define the type a…
parameter estimationIncentive-compatible public transportation fares with random inspection
We consider the problem of designing prices for public transport where payment enforcing is done through random inspection of passengers' tickets as opposed to physically blocking their access. Passengers are fully strat…
BlockingInter-Annotator Agreement in the Wild: Uncovering Its Emerging Roles and Considerations in Real-World Scenarios
Inter-Annotator Agreement (IAA) is commonly used as a measure of label consistency in natural language processing tasks. However, in real-world scenarios, IAA has various roles and implications beyond its traditional usa…
A Survey on Active Learning and Human-in-the-Loop Deep Learning for Medical Image Analysis
Fully automatic deep learning has become the state-of-the-art technique for many tasks including image acquisition, analysis and interpretation, and for the extraction of clinically useful information for computer-aided …
Active LearningDeep LearningDiagnosticMedical Image Analysis