First objective of the study is to find out what causes delay in delivery of procured production materials in an Engineer-To-Order (ETO) manufacturing company. Second objective is to discuss the suitability of existing purchasing related models for eliminating or minimizing those causes in order to improve On-Time Delivery (OTD) performance. Third objective is to find out what are the best ways to organize purchasing in relation to Order Fulfillment Process (OFP) to improve purchasing OTD.

The study provides an overview of market conditions of uncertainty and customization-responsiveness squeeze, manufacturing strategy, OFP and operative purchasing activities. A hierarchical relationship is suggested to exist between them, which is used to explain the impact of purchasing related models of different strategic level to the OTD performance. A brief overview of potential factors affecting the OTD performance is given. Moreover, the study reviews tools suitable for finding ultimate causes of poor OTD performance and general models to improve the OTD performance. A case study of an ETO manufacturing company is conducted and interviews of employees involved in the OFP from different business functions are used as the primary data collection method. Historical delivery records are used as a secondary data collection method to supplement the analysis.

Findings of the study support the view that in order to achieve high delivery reliability, a critical view of manufacturer's own purchasing OFP interaction should be adopted and avoid focusing too exclusively on suppliers' actions. Conclusion of the research is that conceptual frameworks of business process redesign, responsive supply chain, early supplier involvement, information sharing, internal integration, standards and supplier development can be used to improve purchasing OTD performance through tactics of altering product or process design or managing supply or demand. Moreover, it is concluded in the research that a process level approach to purchasing OTD development is necessary because certain item properties are connected with differences in the length or frequency of delivery delays. A framework addressing this finding was developed which can be used to select a suitable purchasing process based on item's properties.