The Roadmap to a Successful Six Sigma Project

There are a lot of reasons that Six Sigma projects fail but they do not have to IF you can stick to the roadmap. I have done lots of projects most very successful but some have failed. In every case we stepped off of the tried and true path to success, the DMAIC roadmap. As simple and easy as these five steps seem to be, you will many times find them difficult to complete. But if that is happening, my advice to you is to “stay on the path”. Don’t skip a single step. If you stay on the path, you will find success.

DMAIC The five step process

So, what is DMAIC methodology? This means Define the issue, Measure the current state, Analyze and identify opportunities, Improve by implementing the best opportunities, and Control the new process to maintain the gains. You start every project at Define working your way through each step until you have put in place Controls to maintain your gains. What many of us do without thinking is we see a problem (Define)and go solve (improve) it. Most of the times you will find that within a year or maybe even a month or week the problem is back. What went wrong? We missed the other steps of the DMAIC roadmap. So let me spend some time talking about each step.

D – Define

The objective of Define is to define the issue (problem) and the real NEED to improve it. I call this need “the burning platform”. It can not be a nice thing to do, it has to be something that will have an impact on the bottom line of the company.

The second part of the define objective is to get alignment and commitment to solve this issue from the project sponsor and the project team. It also includes the team member’s supervision. We need them committed so they will not pull the team member for priorities lower than this project.

M – Measure

The objective of Measure is to go as a team, to where this process is physically and factually understand the existing process. This means collect facts and data not opinions. Everyone has an opinion but few have the facts to back up the opinion. I am not discounting opinions because most folks down in the trenches (and that is where you have to go) are the experts and have excellent idea of what is happening. The thing they lack is the data to prove it. So we listen to them carefully and then collect the data to prove what is happening. Note I said happening, that is not always what the expert says. But with the facts and data we can now go back to the expert and see if they now agree with what we found. Usually they do and are surprised by the findings.

The second part of the Measure objective is to then compile that data you have collected into a characterization of the current state of the process (the baseline for your project). This will show how bad things are or are not. Most of the time things will be worse than they first thought. In some cases, you may find that things are not bad at all. Then you need to explain your results to the sponsor and if the sponsor agrees close the project. You see sometimes even sponsors opinion of what is wrong is not backed by facts and data. So when you collect them it becomes obvious that this was not an issue.

A – Analyze

The objective of Analyze is to take the current state data and analyze it to determine the root causes of the issue. These root causes become opportunities to improve. Measure data shows you the “surface effects” or “pain” the company feels but not usually the deeper root of the issue. Because of that, you will usually find that you need to collect more data related to the measure data that validates the teams opinion of what is causing the current state issue to exist. So here in Analyze we have to take a “Deep Dive” into areas that measure pointed out as really needing improvement.

I – Improve

The objective of Improve is to develop and implement the best plan for improvement of the opportunities (root causes) identified in the Analyze step. There are two key phrases in this objective. “Develop the best plan” and “implement the best plan”. Develop takes some brainstorming and then some experimenting to validate that what you came up with would work. Second in develop is a plan. In the plan you will need several options so that when the time comes for getting an OK to implement it is not one or done (no action taken). Give the sponsor options to choose from but pick your best set and pitch it to them with a why it is best (remember facts and data).

The second key phrase is “implement the best plan. Whatever is picked, you need to create a detailed implementation plan. Create a time line and stick to it.

C – Control

Note: this is the most forgotten step. The objective of Control is to develop and implement the best controls to maintain the gains that the new process is producing. With anything new, things never work perfect. When things go wrong, as they will, you need a plan/ controls that will guide everyone as to what to do. If you do not do this when things go wrong, those involved will revert back to what they know and have done for years. A control plan can be as simple as a log of what happened, or as complex as a statistical control chart. What ever it is it needs to help the people working the new process continue to follow it.

There is a second part to control that has nothing to do with control but has everything to do with recognition. People on and off the team have worked very hard during the project to solve the issue and to keep things going while the team has worked to solve the issue. There needs to be a celebration and rewards for everyone involved to celebrate the success and their contribution to the solution. In today’s business world, we are faster to tell folks what is wrong than what is right so make sure you celebrate your success.

This is just a quick look at the DMAIC process and has not even address questions that should be answered in each step. My plan is to write five more articles each one addressing one of the steps in the DMAIC process in more detail. If you don’t see them at this blog, you will find them at the Six Sigma Knowledge Center.