PREDICTING THE STABILITY OF THE PURCHASE-TO-PAY PROCESS

ABSTRACT In Supply Chain Management, there is still unused potential in the exploitation of data being gathered by companies. An open question is the prediction of supply chain (SC) stability. Supplier and material combinations may lead to constellations where the standard process path is not applicable, leading to problems along the SC. This research develops a technique that helps to ensure stability of SC processes (here: the purchase-to-pay process of a company analyzed for a full fiscal year). Using a data set of more than 100,000 process iterations with 580 suppliers and 34,000 materials, possible predictors of process stability are evaluated whether they have a significant impact and if yes, which is the highest? Therefore, different types of regression are used to extract relevant information. The variance in stability is explained to a certain extent – which is given for each predictor respectively including whether the influence is positive or negative. Having these results, the stability potential for new orders can be predicted, leaving enough time for proactive measures to avoid problems.