Have you ever had an argument with someone close to you (of course the answer is yes!) Did you resolve the issue effectively the first time? The answer is probably no. Typically, these sorts of important, highly emotional problems are very thorny to untangle. Unless these conversations are approached in a rational, formulaic way, the results are often not good! The same is all too often true in business as well.
While most issues in business do not appear to us to be extremely emotional, there is often something more than a profit motive underlying our disagreements. A coworker does not feel respected, someone has an attachment to a project, or perhaps personal issues are crossing the work threshold. Regardless, you work with people and nothing will change that until the robots take all our jobs.
In their book Difficult Conversations, How to Discuss What Matters Most authors Bruce Patton, Douglas Stone, and Sheila Heen describe how we typically fail to address the root of the problem, and what to do about it. Their book is worth the attention of businesspeople for two reason. First, they argue that the methodologies and tips they use are universally applicable, whether talking to a spouse, business partner, or any other close relation. And not only this, but the major premise of the book is that we do not actually talk about the issue when we have these difficult conversations, but that we beat around the bush talking about symptoms of the deeper issues, not the issues themselves.
It is also remarkable how well the methods of these psychologists align with contemporary business knowledge. We’ve spent the last three weeks discussing process design and how to make things work better in your organization. Since of course every organization is going to have to have difficult conversations at some point, it should not be surprising that the psychologists confirm a very specific piece of business knowledge – DMAIC.
Got Problems?
Before we jump into the DMAIC for solving problems, let us first define 3 types of problems:
- Sudden Drop Off – The most commonly identified type of problem, the sudden drop off is characterized by an issue that suddenly occurs after a run of no issues.
- Gradual Decline – These are the problems that sneak up on you. Everything seems to be going fine until you hit a low point and realize that performance has been slowly degrading.
- Performance Gap – A performance gap is when there is a measurable difference between current performance and a desired goal. One could say this is more of an opportunity than a problem, however the methodology and tools presented will still apply.
Once the type of problem is identified, there is a natural progression to identifying the 3 components that are present in the indentation and resolving of all problems:
- Issue – The effect viewed as a pain or symptom (e.g., nonconformance or complaint)
- Cause – The factor(s) that results in the ISSUE (determined by cause and effect analysis)
- Solution – Improvement to address either the ISSUE of CAUSE
Solutions may be selected to either eliminate, reduce, or mitigate the future recurrence of the problem. Very often however, managers address problems on the issue only. Sometimes looking for a quick fix (a.k.a. the band aid fix), or sometime just due to a lack of a formal process. What inevitably happens is that the problem comes back again, and always at the most inopportune time. What is needed is a problem-solving process…
The DMAIC Process
The process of DMAIC is straightforward enough:
- Define – Before we address any problem, we have to define the parameters. What is in bounds and what is out of bounds?
- Measure – If you’re going to meet a goal, you have to have a metric to compare against. Measuring allows you to see if the changes you make have any effect.
- Analyze – This is the bread and butter of DMAIC. Armed with the definition of the problem and a way to quantify it, you can determine the root causes of the issue.
- Improve – Address the cause to resolve issue! At the improvement stage your team brainstorms and implements the necessary changes to counteract the problem.
- Control – Rather than quickly addressing the problem and moving on, DMAIC says that we need to make sure that our improvements are having the desired results and then do something to make them permanent, so the problems cannot recur in the future.
In order to solve any problem, you first have to find the root cause. The Define and Measure functions are aimed at enabling you to identify and learn more about the issue. The Analysis phase is meant to carry out this identification through cause and effect relationships, and the last two steps are aimed at elimination and prevention of recurrence.
Bus Problems Revisited
Consider our bus problems from the discussion of the IPO Model. While the IPO Model was useful for figuring out why there was a problem, we can use DMAIC to uncover and answer the root cause of the problem itself.
With buses, the definition is fairly obvious – they must get where they are going reliably and on time. It is helpful to use a SIPOC diagram to do this, looking at the Suppliers, Inputs, Process, Outputs, and Customers. Notice this is where the IPO Model meets DMAIC, with an S and a C added.
When measuring, the team would brainstorm what data to gather and how to gather it. We would probably look into the total route time, time at each stop, number of riders, etc.
When analyzing, our team will seek out cause and effect relationships by mining the data and applying business expertise. Depending on the level of statistics needed this could involve regression analyses, fishbone diagrams, or any other method of root cause analysis.
Improving is where the rubber hits the road (pun intended) and the team applies their knowledge of the cause and effect relationships to propose and implement solutions, being mindful of the metrics established earlier to track success. Perhaps the solution for our buses could involve changes to the route informed by passenger and traffic data, new policies for riders, or incentives for drivers.
If the improvements are a success, we need to institutionalize them. This is the final stage – control. The team will communicate with management in order to find a permanent solution. When handing the process over to its new owner, the DMAIC team might also provide control charts, performance software, and other means of keeping the process under control.
Only after safely securing the gains can the project be considered a success. Often follow up meetings and additional touchpoints are needed for this. And remember, a process is never done. Processes are always in need of continuous improvement.
We will conclude our process design series next week by discussing robust process design and the DMADV model. The Taguchi method works to ensure customer satisfaction by considering and reducing noise factors in engineering processes.