Should traditional retailers function as pre-warehouses of online retailers?

ABSTRACT

We construct a supply chain to investigate the reason why it is uncommon for the traditional retailer to function as the pre-warehouse of the online retailer. The product only offered by the traditional retailer is labelled as the differentiated product, while the product supplied by both the traditional retailer and the third-party suppliers is called the common product. As the online retailer' common product competes directly with the differentiated product, whether the traditional retailer should supply the common product becomes sophisticated. In this paper, we identify an intrinsic rationale for the traditional retailer not to supply the common product, even if he is unambiguously cost advantageous vis-a-vis the third-party suppliers. This is because the online retailer's order serves as a commitment device that puts the traditional retailer at the disadvantageous follower's position in the differentiated and common products competition; moreover, this quantity commitment allows the online retailer to clinch a better bargaining position on the differentiated product. We further show that it never pays the online retailer to split orders between the traditional retailer and the third-party suppliers. Our results are robust against the traditional retailer's capacity constraint, consumers' preference over differentiated and common products, and supply chain contractual form.

Keywords:

• Differentiated product
• sourcing
• supply chain
• bargaining
• new retail

Acknowledgements

We thank Alexandre Dolgui (Editor-in-Chief), the Associate Editor, and the reviewers for the valuable comments that significantly improved the paper. Hui Xiong acknowledges the financial support by National Natural Science Foundation of China (Grants 71971091 and 71821001). Ying-Ju Chen acknowledges the financial support by NSFC/RGC joint research scheme (N_HKUST615/19). All remaining errors are our own.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 See more details about Amazon Instant Pickup in the following link: https://money.cnn.com/2017/08/15/technology/amazon-instant-pickup/index.html .

2 See the introduction of Missfresh.cn through the following link: https://www.missfresh.cn/gywm/index.html.

3 The strategic cooperation of JD and Walmart is described in the following link: https://www.globenewswire.com/news-release/2017/07/25/1057519/0/en/Walmart-and-JD-com-Expand-Strategic-Cooperation.html.

4 Following Singh and Vives (1984), this demand function is derived from a representative consumer's utility given by $U(q_n,q_s)=a(q_n+q_s)-(q_n^2+2\gamma q_n{q}_s+q_s^2)/2,$ where a>0 and $\gamma \in (0,1)$. Although a common treatment of the derivation of linear demand from quadratic utility in industrial organisation, tacit assumptions under linear demand have been thoroughly studied in literature. Indeed, the strict concavity of consumer's utility in quantities is requested for the derivation of linear demand, as indicated by Amir, Erickson, and Jin (2017). More discussions on this topic can be found in Armstrong and Vickers (2015).

5 In Section 4, we provide an analysis on the situation when the common product is inferior to the differentiated product in quality. It is verified that all the qualitative results hold.

6 In that case, efficiency concern will lead the online retailer to order exclusively from the competitive fringe for the common products. See Corollary 3.1 in Section 3.4 for the results.

7 Our results provide a complete picture regarding the comparison of supply efficiency between a traditional retailer and a third-party supplier. In fact, an exact cost comparison between the two parties could request case-by-case analysis. Issues including the quality gap between the common product and the differentiated product, the traditional retailer's bargaining power in the supply, are highly industry-specific and affecting the efficiency comparison.

8 See more details about the development of e-commerce in China and in the U.S. through the following links: http://images.mofcom.gov.cn/wzs2/202007/20200703162035768.pdf; https://www.emarketer.com/content/us-ecommerce-2020.

9 Source: https://optinmonster.com/online-shopping-statistics/#ecommerce-statistics.

10 Information about this industry is available at http://www.sleeplikethedead.com.

11 Online Supplement A analyses a benchmark case where the traditional retailer sells in off-line stores, therefore does not depend on the online retailer for the selling of his differentiated product. Our central message is unaltered in this alternative structure.

12 In reality the online retailer's market power in profit sharing with the traditional retailer could be much weaker. Online Supplement A illustrates the extreme case when the online retailer has zero market power so that the traditional retailer acquires the whole profit from selling the differentiated products. Our central message is not affected by this setting.

13 When the traditional retailer sells through the online retailer his differentiated product and when they are symmetric in their bargaining powers, $q_c>0$ raises the disagreement profit for the online retailer, resulting in her revenue sharing to be greater than half. While this specific finding illustrates a potential advantage the online retailer can achieve through selling the common product, it is not the driver to our central result. The qualitative result in Theorem 3.1 does not rely on this finding, and will continue to hold when those assumptions are removed, i.e. when retailers are under asymmetric bargaining powers (see the discussion following Theorem 3.1), and when the traditional retailer does not need to sell the differentiated product through the online retailer (see Online Supplement A).

14 It holds that ${\hat{w}}_t>0$ only when $\gamma >0.845$. Thus when $\gamma >0.845$, for w small enough it could be that ${\overline{w}}_t\left(w\right)<{\hat{w}}_t$. If that is the case, in equilibrium the online retailer orders a large quantity $q_c^t$ from the traditional retailer such that v = 1 in their bargaining for profit sharing. In this case, the corner solution ${\overline{w}}_t^{\mathrm{\prime }}\left(w\right)$ is the cutoff value of $w_t$ for the change of the online retailer's sourcing strategy.

15 The intuition is that if the traditional retailer's price is appealing ($w_t$ is relatively low), the online retailer shall order exclusively from the traditional retailer to clinch the Stackelberg leader position. Otherwise, the online retailer shall order exclusively from the competitive fringe to benefit from the low price.

16 Our analysis is conducted under a general assumption that the online retailer has the option of mixing between the two supply channels for the common product. If instead the online retailer has to commit to an exclusive source upfront upon observing the prices of the traditional retailer and the fringe, the finding is unaltered.

17 As shown in Figure 1, the leadership premium ${\overline{w}}_t\left(w\right)-w$ strictly increases in γ. In other words, when the products become more homogeneous, the online retailer's incentive to order from the traditional retailer (and thereby exploit the Stackelberg leader's advantage) becomes stronger.

18 For $\gamma \le 0.888$, we distinguish between two scenarios in the Stackelberg regime. When the fringe is not very competitive ($w\ge w_2$), the optimal wholesale price is set at the ‘interior solution’ $w_t^{\ast }$ as it is effectively unconstrained by the online retailer's incentive problem. When the cost advantage is moderate ($w_1<w\le w_2$), however, the online retailer's threat of turning to the fringe for the common product becomes pivotal. Thus the traditional retailer can only set the wholesale price at the upper bound, given by ${\overline{w}}_t\left(w\right)$, that makes the online retailer just indifferent.

19 Indeed, a meeting between the author and the staff in the JD-Walmart program in summer 2020 reveals that JD and Walmart are no longer placing much emphasis on their integration of inventory.

20 This assumption guarantees that w cannot be too small, so that both the traditional retailer and the online retailer offer positive quantities when the online retailer orders exclusively from the competitive fringe.

21 Assumption B serves two purposes for the scenario when the online retailer orders exclusively from the competitive fringe: first, it guarantees that both the traditional retailer and the online retailer produce positive quantities; second, the revenue sharing proportion of the online retailer never exceeds 1.

22 Assumption C is meant to simplify the analysis and derive analytical results. The qualitative part in Corollary 3.2 continues to hold when alternative values of γ are adopted. For example, if $\gamma =0.6$, the cutoff value of w, below which the online retailer orders exclusively from the competitive fringe, is given by $0.009a-0.017\lambda$; if $\gamma =0.4$, this cutoff value of w is given by $0.003a-0.004\lambda$.

23 When $\gamma >0.888$, in the Cournot regime, $v^C$ increases in γ and $v^C=1$ at $w=0,\gamma =1$. Thus for large γ, the highly disadvantaged profit sharing of the traditional retailer in the Cournot regime leads him to favour the Stackelberg regime, in which he profits through providing the online retailer's common products. As a result, $w_1$ decreases in γ.

24 Parameters are set as a = 10; for the standard case, cost function of the leader and the follower is $TC\left(q\right)=2q$.

Additional information

Funding

Hui Xiong acknowledges the financial support by the National Natural Science Foundation of China [grant numbers 71971091 and 71821001]. Ying-Ju Chen acknowledges the financial support by NSFC/RGC joint research scheme [grant number N_HKUST615/19].

Notes on contributors

Yutian Chen

Yutian Chen completed her Ph.D. in Economics at the State University of New York at Stony Brook in 2007. She is now a professor of Economics at the California State University, Long Beach. Her current research interests include firms' vertical relationship and supply chain management. She has published in journals including International Journal of Industrial Organization, Naval Research Logistics, and Journal of Economics.

Hui Xiong

Hui Xiong completed her Ph.D. in Management in the School of Economics and Management, Tsinghua University and joined the School of Management at Huazhong University of Science and Technology (HUST for short) in 2013. She is a recipient of best young researcher award at HUST and Higher Education Outstanding Scientific Research Output Award in China (Social Science, third prize). Her current research interests lie in operations-marketing interface and supply chain management. She has published in Management Science, Naval Research Logistics, and Decision Analysis.

Ying-Ju Chen

Ying-Ju Chen holds a joint appointment between School of Business and Management (Department of ISOM) and School of Engineering (Department of IEDA) at HKUST. Prior to the current position, he was a faculty in the Department of IEOR at UC Berkeley. He obtained a PhD degree in Operations Management from Stern School of Business at New York University in 2007, and he also holds master's and bachelor's degrees of Electrical Engineering from National Taiwan University. He is a recipient of NYU teaching excellence award, multiple-time Recognition of Excellent Teaching Performance at HKUST Business School, Most Influential Service Operations Paper Award in Production and Operations Management, Harold W. Kuhn Award of Naval Research Logistics, Second place of INFORMS Junior Faculty Interest Group (JFIG) paper competition, Higher Education Outstanding Scientific Research Output Award (Social Science, third prize), Best paper award of CSAMSE (third prize), the Harold MacDowell Award from Stern School, Meritorious Service Awards from Management Science and Manufacturing & Service Operations Management, and other awards and fellowships during his academic journey. He is ranked No. 2 among researchers worldwide by weighted corrected publication rate in Operations Management according to an article in Decision Sciences (2020). He serves as a senior/associate editor for M&SOM, NRL, and POM journals. His current research interests lie in socially responsible operations, operations-marketing interface, and supply chain management. His work has appeared in several leading journals in the fields of economics, electrical engineering, information systems, marketing, and operations research.