How to Choose a Six Sigma Project

Choosing the right Six Sigma project is vital. Several examples illustrate why this is so:

  • If the project is too large or difficult, it will either fail or be cancelled before it can succeed.
  • If the project's scope is too small, upper management may not champion and promote it.
  • If the project's payoff is too small, it will be seen as a wasted effort.

Do we have data?

One important difference between a Six Sigma project and a "normal" project is that in Six Sigma you will use advanced statistical tools.

To do this you will need data. If you already have access to a large data amount it will be possible to dig deep in the "analyze"-step and learn the different tools. If you do not have the data, but have a good idea on how to find it (using Design Of Experiments or other techniques) you will still have the possibility to make a great project, but you will need to plan the "measure" step in detail.

If you do not have any clue on how to find relevant data you should probably try to find another project.


One approach for choosing a Six Sigma project is to start with a brainstorming session. The desired result is a list of potential projects. Then take the interesting step of rating these projects against each other. Finally, present a highly-rated project for executive approval.

The brainstorming session requires having a variety of people, particularly representatives from line functions who will directly benefit from these projects. They should also be able to describe the issue they want addressed, and provide some limits on the scope and timelines.

Select a team of about half a dozen to agree on a fairly small number of criteria, a scoring system, and finally to evaluate the candidate projects. Here is an example of a scoring system that works well for organizations new to Six Sigma:

  • Meets Customer Goals: score 5 if this addresses a specific customer request; 2 if it is known to be a customer-facing issue; and zero otherwise.
  • Reduces Costs, as compared to the cost of the project: score from zero to 5 for no cost benefit, through 1X, 2X, 5X, 10X and finally "above 10 times cost reduction compared to the cost of the Six Sigma project".
  • Improves Quality: Score 5 if this project should reduce defect rejection rates by 1% or more; and zero otherwise.
  • Length of Project: Score 10 for a 2-month project, 3 for a 6-month project, and 1 if the project is likely to take a year or more.
  • Project Staffing: Use the lowest score: 6 if the department can do the project with minimal guidance; 5 if a technical specialist is required; 4 if a Green Belt is required; 3 if a Black Belt is required; 2 if a more senior or structured Six Sigma team is required; 1 if outside Six Sigma consultants are required.
  • Negative Criteria: Each of these adds or subtracts points toward the total for a project.
    • If this issue is already being addressed: subtract 10; zero otherwise.
    • If the solution is already known: subtract 20; zero otherwise.
    • If this project will likely go on forever: subtract 50; add 10 otherwise. (A goal such as "Improve X by Y% annually" cannot be achieved as a project, since it has no end date. "Set a process to improve X by Y% annually" is, however, a project that is completed when the process has been defined and implemented).
    • If the success measures and project drives have already been identified for this issue: add 8; subtract 15 otherwise.

Then take a dozen or so likely projects for closer scrutiny.
It is simpler if each person on the selection team simply contributes their scores to one tally, rather than trying to gain consensus from each person. That does not preclude discussion or debate, especially the discussion to ensure everyone understands the project under discussion.

Remember that these should be objective criteria, requiring research between the brainstorming and selection meetings. For example, estimating the cost to run each Six Sigma project is not trivial; neither is determining the expected cost saving.

Finally, the highest total score should not be the sole determinant of the "best" candidate project.  The CEO may have already stated a goal for, say, improving customer service in a particular division. This exercise may have demonstrated that this particular goal is already being addressed outside of the Six Sigma framework, so then the selection team would advocate for a different project. If the CEO's pet project is a close third from the selection process, however, it might be more important to ensure executive commitment than to pick a slightly higher-scoring project.

By Oskar Olofsson