Purpose

To determine how to best deploy the Triple-A supply chain (SC) capabilities (AAA-agility, adaptability and alignment) to improve competitive advantage (CA) by identifying the Triple-A SC model with the highest CA predictive capability.

Design/methodology/approach

Assessment of in-sample and out-of-sample predictive capacity of Triple-A-CA models (considering AAA as individual constructs) to find which has the highest CA predictive capacity. BIC, BIC-Akaike weights and PLSpredict are used in a multi-country, multi-informant, multi-sector 304 plant sample.

Findings

Greater direct relationship model (DRM) in-sample and out-of-sample CA predictive capacity suggests DRM's greater likelihood of achieving a higher CA predictive capacity than mediated relationship model (MRM). So, DRM can be considered a benchmark for research/practice and the Triple-A SC capabilities as independent levers of performance/CA.

Research limitations/implications

DRM emerges as a reference for analysing how to trigger the three Triple-A SC levers for better performance/CA predictive capacity. Therefore, MRM proposals should be compared to DRM to determine whether their performance is significantly better considering the study's aim.

Practical implications

Results with our sample justify how managers can suitably deploy the Triple-A SC capabilities to improve CA by implementing AAA as independent levers. Single capability deployment does not require levels to be reached in others.

Originality/value

First research considering Triple-A SC capability deployment to better improve performance/CA focusing on model's predictive capability (essential for decision-making), further highlighting the lack of theory and contrasted models for Lee's Triple-A framework.

中文翻译：

---

通过评估 AAA 模型的竞争优势预测能力，寻找部署 Triple-A 能力的合适方法

目的

通过识别具有最高 CA 预测能力的 Triple-A SC 模型，确定如何最好地部署 Triple-A 供应链 (SC) 功能（AAA 敏捷性、适应性和一致性）以提高竞争优势 (CA)。

设计/方法/途径

评估 Triple-A-CA 模型（将 AAA 视为单独的结构）的样本内和样本外预测能力，以找出哪个具有最高的 CA 预测能力。BIC、BIC-Akaike 权重和 PLSpredict 用于多国家、多信息者、多部门 304 植物样本。

发现

更大的直接关系模型 (DRM) 样本内和样本外 CA 预测能力表明 DRM 比中介关系模型 (MRM) 更有可能实现更高的 CA 预测能力。因此，DRM 可被视为研究/实践的基准，而 Triple-A SC 功能可被视为性能/CA 的独立杠杆。

研究局限性/影响

DRM 成为分析如何触发三个 Triple-A SC 杠杆以获得更好的性能/CA 预测能力的参考。因此，应将 MRM 提案与 DRM 进行比较，以确定考虑到研究目标，它们的性能是否明显更好。

实际影响

我们样本的结果证明了管理人员如何通过将 AAA 实施为独立杠杆来适当部署 Triple-A SC 功能来改善 CA。单一能力部署不需要达到其他水平。

原创性/价值

第一项考虑部署 Triple-A SC 能力以更好地提高性能/CA 的研究侧重于模型的预测能力（对决策至关重要），进一步突出了 Lee 的 Triple-A 框架缺乏理论和对比模型。