| This dissertation consists of three essays on bundling in retailing.;Product bundling, why and when. In this first essay, we integrate the existing literature on bundling. We therefore bring together findings from microeconomics and marketing in a comprehensive framework. This framework is based on the insight that bundling can serve different objectives, as a function of demand-side characteristics---viz. consumer preferences for the separate products as well as for the bundle---and supply-side characteristics---viz. cost structure and competitive situation. On the basis of this framework, we point out areas for future research.;Brand performance implications of bundle promotions. This essay concentrates on brand performance effects of constrained multi-item promotions---or bundle promotions---in comparison with un constrained promotions. We distinguish between two types of bundle promotion: "self-bundling"---where consumers themselves compose the bundle by selecting a number of units (the bundle quantity requirement) from a range of items in order to obtain the bundle discount---and fixed-bundling---where the seller determines the composition of the bundle. We unravel the mechanisms through which these promotions affect the brand's net (cumulative over time) sales and revenue. First, we find that unconstrained multi -item promotions lead to a higher immediate and net sales impact than unconstrained single-item promotions. However, multi-item promotions appear to be less efficient than single-item promotions as they tend to generate relatively more forward-buying and subsidization. We then demonstrate that a bundle quantity requirement (self-bundling) can force consumers to buy more and drive up the promotion's efficiency---and thus sales and revenue results---by reducing the relative degree of forward-buying and subsidization. Imposing a composition constraint (fixed-bundling) typically generates a lower immediate sales response than self-bundling, yet, surprisingly, often maintains the low relative degree of forward-buying and subsidization.;Impact of bundle promotions on choice, quantity and incidence decisions. This essay empirically investigates the impact of self-bundling promotions on consumers' purchase decisions for Fast-Moving-Consumer-Goods (FMCG). We calibrate purchase decision models on consumer panel data and promotional information in the potato chips category. We find strong evidence that items that are part of a bundle promotion reinforce each other's choice probability (the leverage effect). Surprisingly, consumers who buy less than the bundle quantity requirement and, hence, do not get the discount, may still switch toward the bundle items (the 'communication' effect). Other consumers switch away from the bundle-promoted items (the 'reference price/betrayal effect'). At the quantity level, we find that the bundle discount induces an acceleration effect that is moderated by the extent to which the bundle pushes that consumer to deviate from her usual purchase amount ('quantity pressure'). Specifically, quantity pressure reinforces the positive impact of the bundle discount up to a critical point, and tempers it beyond that point. Finally, we do not find any significant impact of bundle promotions on incidence.;To further evaluate the magnitude of the promotional bundle effects, we compute average changes in choice probability and expected total purchase quantity for different promotional scenarios. It turns out that---contrary to what one may intuitively expect---a bundle promotion rather attracts heavy users from other brands than it will convert own customers into heavy users. |
