Effective root cause analysis is the heart of any problem solving method such as Six Sigma, Lean, Root Cause Corrective Action and Plan, Do, Check, Act. It is an essential step for identifying critical components of chronic issues, sometimes inside involved projects spanning a few months. It is also the key to daily analysis of sporadic changes in process and product performance in projects you want to complete in a matter of days or hours.
Daily root cause analysis need not be painful and lengthy. Below I will discuss the process of root cause analysis, and some tips to improving the performance of your daily analysis with a minimum of drama and frustration. The root cause analysis steps presented are a subset of the larger DMAIC process of Lean Six Sigma.
Root Cause Analysis From Juran
- Define: Identify a Change in Performance
The entry point to daily analysis is the detection of issues coupled with the clear assignment of the responsibility to act on the issue. Establishment of meaningful scorecards and daily checkpoints are critical to establishing expectations. Create a daily report of key effectiveness measures detailing adherence to customer requirements, and efficiency metrics recording performance compared to production schedules and to material and manpower budgets. Use historical records of expected process variation to set action limits based on special cause variation to avoid overreacting to normal conditions.
To get action on the deviations, teams must be assigned to investigate and correct the problems. The best approach is to assemble a small team drawn from key parts of the organization that have an impact on the problem. Management must clearly assign responsibility to that team and support their investigations and improvements through follow-up meetings, and reinforcement of any improved processes.
- Analyze: Diagnose the Cause
The second step of the root cause analysis process is to obtain symptoms of the problems and identify root causes. It is tempting to avoid this step in favor of conventional wisdom. Do not short-change this step unless the cause is painfully obvious or the risk of implementing an ineffective solution is very low. When at all possible, always back up theories with unbiased data analysis to ensure a clean solution for the problem.
Remember that a symptom is observable evidence of a problem. The detection of a symptom naturally leads to additional questions. When did the problem start? Where do we observe the problem? Where is the problem not happening? Pareto analysis can help narrow the inquiry to the vital few contributors to the problem.
The team uses the symptoms to quickly discover potential root causes of the problem. Simple process flow diagrams and brainstorming may be all that is necessary to identify the range of theories. If there are many theories, the team can organize the potential root causes into a fishbone diagram for an easily understood display.
Finally, test these theories with quick data collection and analysis with simple tools such as summary statistics by stratifications, bar charts, histograms and scatter plots. This step is greatly simplified if you regularly collect data confirming the quality of your processes and products and the efficiency of your operations. Accurate data stored in relational databases supercharge your data analysis when combined with querying tools to quickly confirm or dispel a theory.
- Improve: Remedy the Cause
Now comes the fun part. With the root causes clearly identified, the team can design effective remedies aimed directly at the problem. The team should entertain a range of alternatives, and choose the best ones based on criteria such as impact on the problem, cost, and time to implement. Mistake-proof solutions are clearly the best, as it is a preventive action to keep the problem from recurring.
It is also important at this step to consider organizational culture in the selected improvements. How will the organization be impacted? What will be the response? What should we do to establish buy-in to the new conditions?
- Control: Hold the Gains
It is important to quickly implement the solution once it is identified. However, often the enthusiasm is only maintained to put out the fire at hand, and not to follow up to prevent the problem from coming back. It is critical to set up a control plan to measure the continued performance of the corrected process and to maintain the improvement. This may be simple if listening posts were effectively established in the beginning. These scorecards may have enabled you to identify the problem in the first place.
The circle is complete with the establishment of strong dashboards with daily checkpoints. Each iteration of this process quickens the detection of future issues, and strengthens future root cause analysis and corrective action.
Looking to enhance and develop your skills? Want to learn how to take your organization to the next level? Want to earn a professional certification? Join us for our open enrollment, online Black Belt course starting on September 8th. Limited seats available. Save $$$ if you register with a group of 5 or more! Learn more here. Already know you want in? Register here.
Author: Michael Stamp
Michael Stamp serves as Vice President at Juran. In this capacity, he specializes in Continuous Process Improvement, Lean Management and delivering a variety of quality programs to corporate clients. Mr. Stamp has over 25 years of experience and is an outstanding change agent who can identify opportunities, develop focus and provide strategic and tactical business solutions.
Mr. Stamp’s core competencies include Process Improvement, Operational Streamlining, Data Science, Special Project Management, Training & Coaching, Cost Reduction, Multi-Site Operations, Quality Control/Assurance, Policy & Procedure Development, Leadership Development & Culture Transformation and Statistics.