Following-up on the highly successful webinar with PwC, "Why Elite Procurement Teams Establish Centers Of Analytics" and the subsequent Category Managers Playbook we made available a few months back, I have recently included a simple but highly effective regression analysis model. Few of us are mathematicians but with the tools we have today we are all in the position to easily and effectively bring some of these complex models to bear without having to know much about the engineering but focus on the outputs. It is the outputs their interpretation which is where the value is for us. I have used regression in many different scenarios and found it to be incredibly useful – even for the layman.
I hope the simple model attached enables you to quickly and effectively load the necessary data up and by changing a few data ranges, reach a valuable and detailed understanding of the relationships in your data, thereby giving you a real edge when engaged with suppliers.
What is regression analysis? It is a statistical process that enables you with some accuracy to understand the relationship between such variables as “price and volume”. For sourcing teams this is an important and sometimes complex relationship to model as they structure deals and negotiate.
More specifically, the model helps you understand how the typical value of the dependent variable (ie “price”) changes when one of the independent variables is varied (ie “volume”). You can use more than one independent variable but for simple modelling I have stuck to just one.
I have used regression analysis for prediction and forecasting, where for the geeks out there, there is an overlap with the field of machine learning. For procurement professionals I have seen the benefits such as;
1. Upskilling – Rarely have I seen regression being used in procurement but when it is, it brings a new and exciting dynamic to the analytical process. This simple model can be quickly and effectively brought to bear to support professionals navigate through some of the complexities they face whilst analysing the supply market and establishing the most optimal course of action.
2. RFP and RFI - Once you have collected all your data on volume and price, it would be extremely useful to run this model and see if there is a tight or loose correlation in the supply market and from this decide how best to negotiate and possibly contract.
3. Price Variation Analysis - Much is said about price variation analysis and it is an excellent first step in the process but by including this in next step you take this analysis to a new and more valuable level. By including demand in this, you develop a powerful three dimensional predictive model from which to negotiate.
4. Predict and/or forecast - There is a lot of mystery swirling around this and for most, this first simple step is an excellent way to model and forecast/predict that given certain conditions/volume the average price you will transact at is likely to be close to the line.