- Data Analytics
The procurement organization, which manages up to 80% of total organizational spend, is increasingly recognizing the importance of leveraging data to enhance its processes. A buying channel in the procurement function is the end-to-end series of steps to request, approve, purchase, receive, and pay for goods and services. This capstone project explores the procurement buying channels processes in the sponsor organization, an S&P500 company operating in the pharmaceutical sector, using advanced data analytics to provide insights on variables that influence the use of channels. Our analysis included analyzing the behavior of the stakeholders in the buying channels processes through clustering of the data using a K-Means algorithm. Clusters were added to existing and newly engineered features in the dataset. Machine learning models for regression and classification were then developed, to identify key variables impacting the use of buying channels and to predict buying channel utilization. The results show that there is a confusion in channel usage due to lack of clarity in the classification of channels. Also, predicting the completion time of purchases through the channels had low accuracy due to lack of granular information on transactions. Our recommendations include a new framework articulated over three pillars: enterprise-wide structured and unstructured data integration, a Co-Pilot architecture to support stakeholders through generative AI applications and performance incentives for ongoing learning. Efficiency will be driven by a feedback loop within the buying channels process to incorporate insights gained from each completed transaction to inform future transactions.