Analyse phase criteria for the DMAIC improvement process

The DMAIC Improvement Process comprises five phases namely, Define, Measure, Analyze, Improve, and Control. The Analyze Phase is where the data collected is analyzed to determine root causes of problems being addressed. In this phase we are going to answer the question, “Which variables in the process make a difference, or not?” The key is to identify and validate root causes of issues and determine their impact on the process performance metrics that were established during the Measure Phase. The following article outlines Analyze Phase Objectives, Tools to Apply, and Deliverables to be achieved.

The overall objective of Analyze is to test and evaluate all of the process inputs to determine if changing from their current levels, or settings, will have a positive, negative, or no impact on the process key performance metrics. Process inputs include tools, materials, work instructions, policies, measurements, process settings and controls, software and systems, and even communication methods.

The point is to determine the best set-up for the process so we can predict what the outcomes will be. Analyze Phase aims to determine what causes the variation in the process outcomes. Outcomes, or performance metrics, include cycle time, lead time, processing time, inventory levels, quality of products and services, and customer requirements. The process comprises both Internal and External Customers along with your Suppliers, all of whom have requirements that must be met.


  1. Develop theories or hypotheses about the root causes of the process problems.
  2. Test the hypotheses with data to determine if there is a difference.
  3. Answer the question, “Which variables in the process make a difference, or not?”
  4. Identify and verify the root causes of the process problems.
  5. Determine the best Set-Up for the process that makes all outcomes predictable.

Tools to Apply:

  1. Brainstorming
  2. Affinity Diagrams
  3. Tree Diagrams
  4. Cause & Effect, or Fishbone Diagrams
  5. Process Control Charts
  6. Histograms and Pareto Analysis
  7. Data Collection Plans, Sampling Plans, Check Sheets, and Forms
  8. Detailed Process Maps with Current State Cycle Times, Yield, Processing Times, and Inventory
  9. Process Activity Analysis of Value Adding and Non-Value Adding
  10. Statistical Analyses such as Hypothesis Tests, Correlation Studies, Regression Analysis, and Design of Experiments

Deliverables that Signify Phase Completion:

  1. The complete list of the Identified Potential Causes.
  2. Details about the criteria used to narrow the focus of the causes for further investigation.
  3. Data and Results from Statistical Analysis used to Verify the Root Causes of the issues.
  4. Documentation of how Analyses were Interpreted.
  5. List of Verified Root Causes and How they Impact the Process Performance Metrics.

At the conclusion of Analyze phase, the team has a thorough understanding based upon data analysis of what makes the process tick. Many solutions for improvement have been uncovered and the team is prepared for the Improve Phase where the pilot testing is conducted to validate the achievement of the project goals outlined in the Define Phase.