Let’s look at the main characteristics of each approach:

Kaizen™ – a Japanese word that simply means ‘Change for the better’. The main characteristics are:

• Projects are well defined and baseline stats are collected before starting
• Dedicated resources are subject matter experts (SMEs) and focus on only the project
• The solutions should come from the SMEs as they will need to act as champions for the change
• Often follows the Demming/Shewhart cycle of Plan-Do-Check-Act
• Kaizen™ projects typically last 3 – 5 days
• Management MUST make resources available from support functions during the project. i.e., HR, Finance, Warehouse, Sales, Marketing
• Solutions are implemented based on 80% confidence instead of 95% 
• Implementation is completed within the project timeline but if items fall outside, they are generally completed within 20 days
• Basic analysis is acceptable with indicative results enough to make decisions.

After the Kaizen event the following activities are recommended:

• Walkthroughs are conducted with the process area team
• Follow up with Future State process owner
• Updated Process Metrics, measure, and data
• To-Do list follow up
• The final measure of controlled metrics evaluated
• Training and SOP verification

DMAIC – a 5-step process where the 1st letter of each stage spells out DMAIC: Define, Measure, Analyze, Improve and Control. The main characteristics are:

• The existing process is not meeting customer requirements, but the reason why is not obvious
• Time is spent on analyzing the baseline data to understand current performance
• Baseline data is used to prove/validate the benefits once re-measured
• Solutions can come from anywhere and may not be popular with employees as they may mean significant changes
• Solutions require 95% confidence in being correct before implementation
• There can be a level of risk associated with the solution that will need to be accepted by the business before implementation
• Change is led by a Six Sigma Blackbelt or Greenbelt due to the nature of the data analysis.