The first step in the "Multi-Criteria Customer Ranking" (which I also call Customer Optimization) process is to determine the sales and operations/supply chain metrics that are important to your company. This Supply Velocity Client both produced and delivered most of their products. As you can see on the link below (Slides 2 and 3), this company balanced their key performance metrics amongst growth, productivity and cash flow. In addition to the customer profit mining project, they now use these scorecards as part of their monthly business review. (Note, the red-yellow-green performance targets are blacked-out for confidentiality purposes.)

http://www.supplyvelocity.com/files/Mining_Customer_Data_for_Profitable_Growth.pdf  

The next step is to determine the metrics we will use in the Multi-Criteria Customer Rank. This is critical because it is these metrics that will guide us to finding where we have high-revenue but low performing customers. The difference between a high-revenue and low Multi-Criteria rank is where we will discover how to increase profit from your existing customer base.

We began with a list (Slide #4), which comes from our experience, academic literature review and input from a panel of company CEOs. This Client added a few that were specific to their business.

From the 19 metrics on Slide #4, the team chose 5 metrics to define a customer performance; annual gross profit $, pay quickly (as measured by average days to pay invoices), average revenue per delivered order, year over year growth % in revenue and gross margin %.

The 5 metrics were then weighted based on importance in creating the customers multi-criteria score (later referred to as the Overall Performance Factor, or OPF). This is shown in Slide # 7. What you see are 4 rounds of voting by the team. There were three members of the team, the CEO, VP of Sales and VP of Operations. This method is called the Delphi Panel, which was originally used by economists to create a consensus on economic measures. It turns out to be a very handy way for business managers to also develop consensus.

Finally we put these criteria into a data transformation formula. This is shown on Slide # 8. I cannot show the specific customer data (it is Client-confidential) so we are showing the normalized (on a 0 - 1000) scale. Normalizing the data is part of this calculation. The customers you are looking at are the top 20 revenue-rank customers.

Finally, we get to the best part of this analysis. We compare the Revenue-Rank of each customer to the Multi-Criteria Rank. This list of the greatest delta is shown on Slide #9. We will pick out one particular customer, Ar64, for more detailed discussion. On Slide #10 we show this customer compared to the average for all customers. The improvement opportunity jumps out. First, we see that the gross profit $ and growth % are above average. This indicates that customer Ar64 is an above average revenue customer and is growing. In fact they are the 37^th (out of about 350) revenue-ranked customer. However, they pay slowly, are ordering too little with each order and are ordering low profit margin items. The Sales Representative needs to improve their payment terms to about average; which will provide an immediate cash flow benefit. If she can increase their order per delivery this will greatly improve or logistics productivity, reducing the cost-to-serve this customer. Finally, they order lower profit margin items. The Sales Representative needs to present higher margin items to this customer, which can also increase the average size of each order.

The implementation of Customer Optimization manifests itself in each customer with a negative Revenue-Rank to Multi-Criteria Rank, getting a customer performance improvement plan. This is real data to make real, measurable improvements in customer performance.

We realize, because we did not show you actual customer data that it can be hard to follow all the numbers.  We didn't to maintain Client-confidentiality. If you would like to learn more about Multi-Criteria Customer Ranking and Mining Customer Data for Profitable Growth method we would be happy to explain the model in greater detail. Just send an email or call (314-406-4962) our President, Mitch Millstein.