Spend Matters welcomes this guest post from Maulick Dave, of GEP.
One of the most popular KPIs used to measure the performance of a procurement organization is savings realized. However, the magnitude of the savings projected at the goal setting phase for the year, or at the start of a sourcing project, is often looked upon with skepticism. Has my CPO given me too aggressive a savings target for my category or has my category manager understated the savings target to manage the downside and delight me later? How can you verify such cases and set a practical savings target that will improve your planning and decision-making abilities?
Welcome to the world of predictive analytics! Procurement is not new to the world of analytics. Many organizations have been using analytics for spend profiling, supplier performance evaluation, compliance management and many other similar use cases. However, a strong mathematical foundation for computing the savings target was always elusive. I’d like to introduce what I have dubbed as the Savings Regression Analysis (SRA) model, wherein we use multivariable regression (a statistical method to determine the relationship among various variables) on a dataset of past realized savings for a given subcategory to compute the savings potential under current market conditions.
An outcome of a sourcing activity largely depends on the criticality of the commodity being sourced and complexity of the supplier market within your organization or industry. The criticality determines how much risk an organization is willing to take on with respect to souring of the commodity, and the complexity of the supplier market determines on whose side the balance of the power is likely to be titled — supplier or buyer. The SRA model quantifies the above two parameters across a number of explanatory variables (drivers of savings) to predict the response variable (savings) in different scenarios using statistical models.
To build an SRA model, the data needed could come from the organization itself or from the industry. The more data you have, the more accurate the results you are likely to get. At GEP, we rely on our trusted internal knowledge bank for most of our data requirements. Typical explanatory variables would be: dollar spend, unit volume, buyer’s market share in the industry, criticality of the commodity (high, medium, low), last competitively sourced and for the supplier market, the number of eligible suppliers, market growth, distribution of the market share and other category specific variables.
Note: Your explanatory variables will vary from category to category and you are free to choose variables that you think drive savings. However, remember more variables do not mean better results. There is an amount of statistics knowledge required to determine the optimal number of explanatory variables. In running a regression, you will get an equation showing relationships between savings realized and your explanatory variables. Fill out the equation with values that reflect your current market conditions and voilà! You have your savings estimate. You can also run different hypotheses and get a confidence interval on your savings.
Advantages of SRA
- Better cash flow management: You have a strong analytical foundation to say there is a Z% chance that you will have your savings in the range of X%–Y%, which will help in your working capital management
- Opportunity to time the market: Adverse supplier market conditions give you an opportunity to extract maximum value for your procurement activity. SRA helps you make the decision of when is the best time to go for your sourcing activity
- Goal setting: SRA helps you set the expectation right. You don’t have to look for a face-saver at the end of the year because the target you set at the start was too high
- Performance management: It will be easier for management to figure out which department has beaten the curve and delivered exceptional performance.
For more interesting thinking on procurement, visit the GEP Knowledge Bank.