Section 7.7. Improve

7.7. Improve

The I in DMAIC is for improvement. Through the collection and analysis of data, the Six Sigma team will reach a point where it has amassed enough information to determine how to make the process better. And so, it then works to improve the process, making changes so the process is more effective, more efficient, or both. All of this comes from the insights provided by the data. With Six Sigma, the solution is always in the data.

Through data collection and analysis, you will come to understand how the process performs, what its strong points are, what its weak points are, what kinds of defects are entering into the system, and where specifically they are coming in.

You can think of the analysis phase of Six Sigma as the point at which you identify the root causes of defects. You can think of the improvement phase as the point at which you remove the root causes.

The improvement phase generally consists of a series of six steps:

  1. Assess

  2. Develop

  3. Select

  4. Modify

  5. Pilot

  6. Verify

Let's take a brief look at each.

7.7.1. Assess

Through the analysis of the data, you and your team will have at hand a collection of improvement indicators. These indicators will point to where the process has significant variation, where it tends to become more unpredictable than at other points, and where it begins to drop its level of control. This is where you and your team should begin to assess your options. You focus on the root causes of defect introduction. The data should give you multiple ideas as to how to modify the work stream. Some of the options will hold more promise than others. Some may appear to hold promise but prove for other reasons to be impractical. Some may appear quick and easy to implement but may not show strong potential for change. Six Sigma is founded in data, but here you should bring your professional judgment into play, and that of your team, too. Assess the options before you. Review them against the backdrop of the organizational culture, the budget for process improvement, the schedule you are working on, and the goals you have established in your project plan.

Once you have critiqued the options, you can begin to develop some of the promising ones.

7.7.2. Develop

There are a few paths you can take when you begin to develop potential process improvement solutions. The analyses might indicate that one path might be to make corrections to existing process components. This may be the most expeditious route. If you can fine-tune process elements that already exist, you may be able to enhance performance without having to perform a lot of re-engineering. In mature process systems or in highly developed systems, this is usually the kind of action taken.

Another path may be to create new work flow extensions to account for missing process components. This comes about through data that shows root causes are due to unaccounted-for activities, the absence of which allows for significant performance variance. These kinds of situations are usually seen in relatively new process systems or in systems that work in dynamic environments where evolutionary change is continually in effect.

A third path may be to define new processes or new components to reroute work flows toward greater efficiencies. This path naturally requires the greatest amount of work. But it does deliver the advantage of allowing for the greatest degree of impact and the greatest degree of control.

The idea is to develop some of these ideas to the extent that their benefits become apparent to you and your team. Then you can make a decision to take the one with the greatest strategic potential and develop it for deployment.

7.7.3. Select

Here's what you've done so far. You have analyzed the data. You have assessed it for solution potentials. You have focused on certain possibilities and have begun to develop them to ascertain their promise for positive impact. Now you and your team can begin to focus on the hard solution. Study the options you have developed and make a prudent judgment to focus on the best one for the project. Select the solution to develop.

Keep the concept of practicality in mind. The solution with the most potential for change may not be the most practical at this point in time. Think through the needs of the organization, the proportions of your project, and the capabilities of your team. Then choose the solution that you know you can successfully implement while obtaining as much of your goal as you can.

7.7.4. Modify

Now you improve the process by eliminating the root causes of defects. You improve the targeted process by designing creative solutions to fix and prevent problems. You can also explore innovations and improvements using relevant technologies and improvement disciplines. The modifications should reflect the trends and constraints reflected in the data. In fact this modification step should be one of the simpler ones in this six-step activity. The only temptation here may be to veer creatively from what the data tells you. Try to resist that. Stay close to the strategic direction you have established so far, and the improvements should work well.

7.7.5. Pilot and Verify

Once you have the revisions down or the new elements in place, you should try to run the process through a real-life-like situationa pilot. Piloting the solution is helpful for a number of reasons. You want to test the pilot out in an environment as close to production as possible. Test the working of the process and then evaluate the results of the pilot.

Once the pilot is completed and you and your team are fairly certain that the results recommend deployment, you should verify this conclusion with the rest of the organization. Up to this point, the Six Sigma effort has been largely confined to your team. Now, it's important to bring the organization in more fully on your efforts. Even though you may have full authority to deploy any improvements you deem appropriate, you should present your case to the organization in a formal manner.

The first step here is to thoroughly document the proposed improvement. This includes describing the existing process and its performance metrics, describing the proposed enhancements and the performance metrics gained from the pilot, defining the benefits to be realized from deployment of the proposed change, and outlining the kinds of deployment activities that may be required to install the improvements in the existing system.

Once the documentation is complete, you can present it to the organization. This is typically done as a peer-review session. You work with the organization to identify and select those knowledgeable people in the system who will be impacted by the change. They should have a chance to review the documentation, ask questions, and comment on it. The objective is to get the approval of this group, to gain their consensus that the deployment should proceed.

Figure 7-10 illustrates the improve phase of DMAIC.

Figure 7-10. In the improve phase of DMAIC, the objective is to use the data analyses to formulate a path for improvements, and then take actions to make the improvements. Here you will use the data results to develop potential solutions, evaluate which are the most promising, and then refine the process to reflect the new approach, making it ready for deployment into the organization.

7.7.6. "I" Is also for Implement

The "I" in DMAIC is for "improve," but you can also think of it as standing in part for "implement." We'll see in the next sectionin "C"that the deployment of the improved process can be handled through a Control Plan. But sometimes the Control Plan has more to do with maintaining the performance of the process in the production environment. Where the idea is placed is not as important is that the concepts are carried out. The focus of implementing is to carry the improvement out into the organization, to deploy the solution in a coordinated and organized manner.

Some of the basic tasks here include developing a deployment/implementation plan, socializing this plan with the organization to gain support and cooperation, and then deploying the improvements according to the plan.

7.7.7. Tools for the Improve Phase

Some common statistical tools used in the improvement phase include:

  • Design of experiments

  • Analysis of variance

  • Multivariate regression testing

  • Process simulations

  • Failure mode effect analysis

Quick Take


The improve phase includes the activities that result in process improvements. Here, based on the analysis of the data you've collected, you focus on potential improvement points to increase efficiencies and remove root causes of defects.


Examine the analytic results from the data and assess potential opportunities to make improvements.


Based on the opportunities, develop alternative solutions, documenting the rationale as to how each might result in improved performance.


Evaluate each potential solution and then select the one that appears to promise the greatest return for the organization.


Modify the existing process in line with the structure of the solution. You'll emerge with a new refined process.


Where practical, pilot the new process in a production-like environment; evaluate the changed performance.


Measure the performance of the process in pilot and verify that it operates as planned. If it does not, you may take it back to redesign. If it does, you can plan to roll it out to the organization at large.

Process Improvement Essentials
Process Improvement Essentials: CMMI, Six SIGMA, and ISO 9001
ISBN: 0596102178
EAN: 2147483647
Year: 2006
Pages: 116 © 2008-2017.
If you may any questions please contact us: