Customization management in supplier performance

Prior research on customized component transactions asserts that from a manufacturer’s perspective, customization costs can be reduced by creating collaborative relationships. However, there are few researches on the supplier’s perspective. In this paper, a survey of Japanese suppliers revealed that (a) supplier performance improves when there are more proposals both from and to customers, and that (b) supplier performance deteriorates when proposals only come from the customer. In other words, in case of the top-down relationship in (b), supplier performance deteriorates, but in the bidirectional relationship in (a), supplier performance improves.


Introduction
Customized components are parts made according to a customer's (or manufacturer's) specifications (Asanuma, 1997). The opposite of this is an "off-the-shelf product," which means the parts offered by catalog. Collaborative relationships between manufacturers and suppliers are found mainly in Japan's automotive industry, where they facilitate customized component transactions by, for example, reducing transaction costs (Dyer & Chu, 2003) or the length of the development stage (Clark & Fujimoto, 1991). In other words, Japanese manufacturers gained a competitive advantage globally because many of their suppliers have been able to make high-quality customized components at low costs in accordance with the customers' (manufacturers') needs.
However, these studies have usually focused on analyzing manufacturer performance, with no regard to the performance of the supplier. An exception is Stump, Athaide, and Joshi (2002), who asserted that customized component transactions increase sales potential, so that suppliers actually gain a competitive advantage when they meet manufacturers' customization needs. However, few studies have focused on the impact of the customization of the supplier's performance (Wang, Lee, Fang, & Ma, 2017).
In fact, customization costs will reduce suppliers' profits. To prevent this reduction, engaging in close communication based on collaborative relationships between manufacturers and suppliers will promote information-sharing and help to solve problems (Takeishi, 2001). In addition, information-sharing between manufacturers and suppliers, such as the purpose and functions of customized component, could reduce customization costs (Nakagawa & Song, 2016). Moreover, prior research indicates collaboration with manufacturers, for example, information-sharing and problem-solving improves supplier performance and broadens customer scope (Konno, 2003(Konno, , 2007. However, prior research on information-sharing has paid little attention to differences of who sends the information. The impact on supplier performance is considered to differ depending on whether (a) the proposal was made by the manufacturer to the supplier (proposal from customer) or (b) by the supplier to the manufacturer (proposal to customer). In fact, the prior research on business solutions company emphasizes that high profits are to be made by providing new value to customers (Nobeoka, 2011), andKuwashima (2004) showed that in the chemical industry, development is successful when proposals are made to the customer before the customer specifies what they want, rather than simply responding to whatever the customer says. Therefore, this paper conducted a quantitative analysis of suppliers to several industries in Japan with respect to the impact on supplier performance of (a) the proposal from customer and (b) the proposal to customer. The IMSS surveys companies in terms of their current sales and profits compared with the sales and profits three years ago, using a five-point scale ranging from "much lower" (1) to "much higher"

Method
(5). Sales and return on sales are the dependent variables in this analysis.
The independent variables are the following three: (A) Degree of customization (DC): In these questionnaire items, customers were asked to characterize their transactions with three types of customers-(i) top tier, (ii) second tier, and (iii) mid-or lower-tier 1 -on a five-point scale ranging from (1) pure standardization to (5)  were asked about (i) top tier, (ii) second tier, and (iii) mid-or lower-tier, respectively, and because of the strong correlation, we used the aggregate for the three types as a synthesis variable.
(C) Frequency of customers' proposals for product designs to the supplier: For this variable as well, respondents were asked about (i) top tier, (ii) second tier, and (iii) mid-or lower-tier, respectively, and because of the high correlation, the three responses were aggregated and used as a synthesis variable.

Results
We divided the three independent variables of A, B, and C into two groups-those above the mean and those below the mean-to conduct our analysis.

First, we created cross tables for (B) proposals to customers and (C)
proposals from customers by splitting (A) into high and low groups. The low DC group is shown in Table 1, and the high DC group is in Table 2. The low DC group makes products that are almost off-the-shelf goods, so the distribution was mainly (53%, or 20/38) in the cell in which both proposals to customers and proposals from customers were low. Meanwhile, for the high DC group, there was a significant positive correlation between proposals to customers and proposals from customers.
However, even in the low DC group with almost standardized products, suppliers with a high number of (B) proposals to customers were performing significantly well in terms of both sales (F(1, 34) = 3.651, p < 0.1) (   (Table 6). In addition, interaction was also significant for return on sales (F(1, 31) = 3.242, p < 0.1). In other words, as shown in Figure 1, when the frequency of proposals to customers was high in the high DC group, return on sales was high when the frequency of proposals from customers was high. However, when the frequency of proposals to customers was low, return on sales declined when the frequency of proposals from customers was high.   High proposal to customer Low proposal to customer

Conclusion
As summarized in Table 7, return on sales tended to be lower for the high DC group than that for the low DC group. 2 However, as is clear from Figure 1, the problem is when the frequency of (B) proposals to customers is low and the frequency of (C) proposals from customers is high. In other words, when suppliers have customization requests from customers, without a proposal to customers, return on sales is greatly reduced. However, as can be seen in Figure 1, when the frequency of (B) proposals to customers is high, return on sales is high when the frequency of (C) proposals from customers is also high. Thus, in the high DC group, when the frequency of (C) proposals from customers is high, suppliers may see their return on sales suffer unless they become proactive in making their own proposals.
It has been noted that because size of Japanese suppliers are smaller than manufacturers, it is structurally difficult for suppliers to turn down customization requests from manufacturers (Takashima, 1998