Process Improvement And Design of Experiments (DOE)


    When you consider a process for improvement, there are a number of things you could do it - Train the people better, give them more latitude in decision making and keeping customers happy who bring in more business, invest more in a rules engine that automates many of the otherwise manual decision making, buy three additional servers so that system response time is better or add additional people.

    All of these cost DIFFERENT amounts of money, effort and dedication and may produce results of different magnitudes! What is interesting is that some of the results may be different from what was expected and may come as a total surprise! May be adding some better lighting to the location where work was done produces an improvement much more than millions of additional dollars spent on a new software application! The only way to find out is to experiment in a limited way first before committing one or more courses of action among the many available.

    This is where Design of Experiments (DOE), originally from the natural and social sciences and now being adapted for use by Lean and Six Sigma experts comes in handy as a rational way to go about process improvement. Every organization has its limits when it comes to resources and so money judiciously spent is always much more appreciated than money thrown wildly at the problem without having some idea of what the end results might be for a dollar spent on each of the different options!

    Design of Experiments advocates that perform experiements on a smaller scale, measure the results, compare them to your expectations and use the conclusions further in whatever way your overall objectives dictate.

    Sometime ago I had written an article about Cause and Effect Diagrams and Design of Experiments in Process Improvement. These two tools are very useful in effecting rational, focused process improvement that makes the best use of a dollar of resources you may have available for you. The Cause of Effect diagrams systematically may help you identify all the candidate causes for the process Key Performance Indicator (KPI) you are trying to improve. For example, in a Call Center you are trying to improve the Average Handle Time of agents on the phone trying to support customers. The AHT metric can be improved by addressing Additional Training for Agents, Upgrading to a new software application or restructuring your incentive packages. Each of these may have different results in the end and may cost you different amounts of money. How do you ensure that you Optimize your improvement? By actually conducting experiments and measuring the results for each of these approaches. Send a smaller group of agents for training, ask the software vendor for a limited set of licenses for a subset of agents to use the new software application in a pilot implementation  or restructure the incentive package only for one team of agents. Measure the results, compare the costs of each and you will find the best ways to spend your limited amount of money for optimum results!

    DOE and Cause of Effect Diagrams provide a useful way of making rational process improvement decisions!

    I think that in the discussion of natural problems we ought to begin not with the Scriptures, but with experiments, and demonstrations. – Galileo Galilei


    Please enter your comment!
    Please enter your name here