Daily Archives: July 2, 2006

Problem Solving Series III: Find the Root Cause

This is the third post in a series of posts designed to introduce you to problem solving strategies that you can use to attack your sourcing and supply chain problems. Last Sunday we discussed three strategies that you could use to help you simplify the task, which is often the best methodology you can use when trying to formulate a difficult problem. Today we are going to discuss five strategies for identifying the root cause of a problem, which is one of the keys to model construction, the third step of the operations research modeling process.

( 1 ) Define before and after

Problems arise as the effect of an observable action and are easily identifiable by their after effects.

First example: your need to reduce spend on a core category.
Why? The procurement plan that was implemented was inappropriate for the category.
Effect: your spend for the period was higher then it should have been since not all expected savings were captured.
Fix: Find a way to improve the process so that all negotiated and identified savings materialize.

Second example: you use too many suppliers for basic commodities.
Why? A proper supply base was not designed up front.
Effect: you are using dozens of suppliers where only three or four would do.
Fix: put a plan in place that identifies preferred suppliers, contracts at best rates, and make sure mechanisms exist to enforce it.

( 2 ) Organize information into a table, chart, or list and look for patterns

For example, lets say total cost of ownership on one commodity in a category is calculated as significantly higher per unit then it should be. Create a table that breaks out the cost components and compares them, as percentages of unit spend, to the other commodities in the category. If, for example, the average freight on the other commodities is 10% and you see freight for this commodity is 15%, chances are this is (one of) your cause(s). Without this comparison, you might not be able to easily identify where the excess spend is.

( 3 ) Try to make the problem worse

This might sound ridiculous, but it will help you validate that you understand the problem. Let’s say your purchase order process is taking too long. If you fully understand the process, you should be able to add an extra step that lengthens the process by a predictable amount. Then you can be sure you understand the process appropriately and that the technology you ultimately select to automate it will be the right one.

( 4 ) Compare situations without the problem

For example, let’s say one supplier is consistently late with shipments. To try and understand why, look at suppliers who are consistently on time. Are communications being handled the same? Are they using the same systems or different systems? Is it potentially a problem with the 3rd party carrier?

( 5 ) Consider multiple causes and interactions

Not all problems will have a single cause and a single solution. In the case of our consistently late supplier, it might be due to delays in supplier acknowledgement and delays in third party carrier deliveries. Look for all potential sources to make sure your fix is the best one.

Next Sunday we will discuss some strategies that will help you in the fourth step of the operations research modeling process, finding a solution.