Communicating the Goals of Collecting Information

Communicating the goals and methods of any data collecting effort is important because this sets the stage for all subsequent events and efforts. Collecting the wrong data will confound our improvement effort. For example, consider an airline that wants to understand the actual arrival of the aircraft against the expected or documented arrival time. Which is better, to measure the time the aircraft gets to the gate? Or is it better to set the time of arrival as the aircraft touches down for landing? The customers will likely call on time arrival something different than wheels down at the destination airport. This discrepancy will be pronounced on those landing where the aircraft sits on the tarmac due to traffic for several minutes causing missing connecting flights. The goals or objectives are important in selecting the attribute or data measured. It is important to collect the right or correct measurement or data. It is equally important to collect the measurement correctly.

Collecting data wrong will lead us to conclusions that may not be true. Collecting data wrong can be found in the data selection methods: what parameters are valid, what data is valid. For example, we collect specific process data, but the event for which we are collecting that data did not adhere to the defined process. Does this data mean anything when it comes to deciding about the process?

There are also unintended consequences possible when we do not articulate the goals and the use for the data we are collecting. This next problem is a combination of corporate culture and lack of transparency in the data gathering objectives and is best demonstrated in a story. Consider a fast food restaurant that is collecting data regarding the speed at which the customer order is made and subsequently concluded. Cars come through the drive-through and then move to the paying window. Upon paying, the vehicle is asked to move forward, away from the window, even though there are no cars behind; there is no line. We are given the illusion that the customer’s order has been quickly dispatched, but the customer is still waiting. The measurements from this event, or similar events, then skew the reality of the data.

How to Analyze through Root Cause Analysis

This is not a book about root cause analysis, however there is much to learn from root cause analysis, and it is possible to come to errant conclusions when performed poorly. Tlie root cause analysis effort is used to correct results from the various parts of the organization. These explorations are initiated from some performance malady or desired improvement area. The formal exploration into the root cause is a mechanism for team members to learn, along with improvements in the organization. Formalizing root cause analysis ensure some measure of rigor and repeatability, or repeatable outcome. Approaches to root cause are important, we do not want to randomly associate cause with effect. Root cause analysis is not near as easy as some seem to believe. If you really want to see what complex root cause analysis really looks like, check out the Smithsonian Channels Aircraft Disasters. This provides a glimpse into what it takes to truly arrive at a conclusion that can help us improve the organization. Errant root cause analysis will lead us to wrong conclusions, then corrective actions that do not mean anything, then improvements that do not improve anything at all. This will be a waste and more importantly at the end of the effort we may walk away believing we actually improved the situation. At the end of this, eventually, the problem will once again rear its ugly head, leaving us to go once again down this path, repeatedly until we accidentally hit upon the root cause.

Informal root cause process provides no mechanism for distribution, neither does it ensure a repeatable or consistent outcome. This inconsistent outcome and lack of distribution means an informal process does not readily support effective learning.

Why Formalize Root Cause Analysis

There are many approaches to determining the root of our problems. In the automotive world, there are two typical approaches, the 8D or the 8 Discipline, or the A3 (named for the paper size). There are some benefits to a formalized root cause analysis. We can think of 6 reasons as listed below:

■ Controls jumping to conclusions and just working on the symptom

■ Repeatable

■ Coordinates effort / focus

■ Documents actions so we know what have tried and what is next

■ Traceable for future events

■ Provides a mechanism by which learning is shared

Formalized approaches can help control jumping to conclusions and just working on the symptom. A formalized approach to root cause analysis is such a way that will keep us from jumping to conclusions about the nature of the failure. We are a big fan of aircraft disasters. If you really want to understand what solving complex problems is really like, check out that show. It is seldom that the first thing we think is the problem is in fact the problem. There is also a significant chance that the root cause is not in fact a single thing, and very likely not the single thing we may immediately believe to be the problem often based upon our biases and experiences.

Root cause anslysis and post-project reviews are material from which to learn and distribute

Figure 3.17 Root cause anslysis and post-project reviews are material from which to learn and distribute.

Formalized approaches, if we have done the work well enough, will generate notes, tests, and results and other documentation that would make it possible to recreate our exploration in as much as it could possibly be replicated. This gives us a sense of really understanding the nature of this problem. If we can replicate the problem, we are theoretically better able to solve the problem, or more importantly, we are able to see if we have solved the problem. The solution should prevent the problem from recurring.

Formalized approaches facilitate coordination of effort from the diverse perspectives, facilitating the nature of the problem and methods for us to explore these beliefs. With a team, we may get a multitude of ideas as to the root cause, and without some level of formalism, we may all charge off in many directions pursuing our own agenda or thoughts on the matter. Our goal is to quickly understand and solve the problem. With team members going off in many and unknown or uncoordinated directions, we are not efficiently making use of the organization’s resources, time, and talent. A formalized approach helps keep the team connected to each other and to the objective and ensures the opportunity for team learning.

Formalized approaches document actions so we know what have tried and what is next. Without a formalized approach we would not know what things have been explored, what has been discovered, and what has been learned, nor be able to distribute that learning. We are able to build upon what has been tried and what has been learned. We can consider our next actions in the context of our past actions in the course of resolving this problem.

Formalized approaches provide a traceable foundation for future events. The formalized process will have some measure of recording the actions that are undertaken and the results of the exploration and the final solution. This information can be fodder or future work. We can capture this information in a searchable database that will make recovery of these activities possible rather than accidental by other people in the organization.

The 5 Whys

Another way for the team to understand how things are connected, including any anomalous performance exploration, is to use the 5 whys. This is exactly like it sounds, ask why and after every answer, ask why again. This ultimately leads us to the point where the root cause of the anomalous performance, stopping at the first why, leads to treating only the symptom and not the source of the problem. In fact, experience suggests often the true root of the problem is not addressed, but what amounts to a band aid rather than fixing the problem in a way that eliminates or

АЗ is an approach to determining the root casue

Figure 3.18 АЗ is an approach to determining the root casue.

greatly reduces the probability of the event recurrence. In some cases, we may have to go even further than the five whys to really get to the source of the problem.

Don’t play the blame game; you should be asking “why” five times, not “who.” Don’t stop the inquiries when you reach a “who” because this usually means somebody will get accused of causing the problem. Blaming does not help with psychological safety, and help with free communication, as well as facilitate experimentation, by eliminating the fear of reprisals,

Be mindful that there could be several valid reasons for a problem happening. Unfortunately, this method usually proceeds along just one path, which might not lead to one of these reasons. Even if it leads to a real cause of variation, this may not be a major cause, much less the largest one. In addition, the five-whys approach is not inclined to lead you to an unlikely or unknown cause, which is often the type responsible for a chronic quality problem. That’s why this method is better suited for identifying the source of assignable variation where there is usually just one cause.[1]


One form of root cause analysis is the A3, which is a graphical representation of the effort to determine the source of the anomalous or failing event. One of the benefits of the A3 approach, is that it encourages graphical representation of the problems and also the solutions and corrective actions. Visual representations of the problem as well as the solutions. A single page of information that describes the situation and the exploration and progress we are making in this regard.

  • 1. Identify the problem
  • 2. Visual representation of the problem
  • 3- Set the new target
  • 4. Determine the root cause
  • 5. Develop countermeasures
  • 6. Institute countermeasures
  • 7- Measure performance - is this better?
  • 8. Standardize, make this common practice

Each root cause the A3 is used to solve, represents some measure of learning by the team. Therefore each of the A3’s are the repository for this learning, albeit not necessarily the best solution as a repository of the resulting work. This is not a searchable content but may end up in a book of all of the A3’s similar to the lessons learned from project management that is sometimes included in a notebook. There are other was to store this information rather than a myriad of seperate documents that must be poured over to understand what may have been learned.

The 8D is another method for exploring and discovering the root of the problem and corrective action

Figure 3.19 The 8D is another method for exploring and discovering the root of the problem and corrective action.


The 8D, also known as the 8 Disciplines, is another automotive approach to root cause analysis. It is called the 8D, which stands for 8 discipline consisting of 8 steps or phases. This sequence of events lead team through the discovery and experimentation work with just enough formality to provide the structure of steps. This includes the close out portion of the work where we thank and congratulate the team.

  • 1. Plan
  • 2. Team
  • 3- Interim plan
  • 4. Root cause
  • 5. Choose and verify corrective action
  • 6. Implement and validate corrective actions
  • 7- Preventative measures
  • 8. Congratulate the team

Like the A3, it is one thing to do the work, learn and solve the problems origins and corrective actions, it is another to put this information in a format that will be easy for those that are not part of this particular learning events, to at least become aware of the existence of these documents and learning.

Quantitate Analysis through Process Metrics

Where we have processes that are actually adhered, we are able to make some quantitative assessments of those processes. Control charts over time, are examples of this data collection. The sum of all of the specific process measurements provide us with information on a larger scale from which we can make decisions. This is especially true for strategic planning. Knowledge of the present performance via process statistics and understanding how these impacts the strategy

Unearthed! Finding Golden Nuggets of Knowledge within Performance Analytics

A review of these process metrics with a view of viewing how the performance and capabilities change over time in a dashboard type view, makes it possible to truly discover the strengths of the company upon which other things can be built. Understanding what works well or the strength of the company, are like discovering gold nuggets. These are the areas that can propel the organization to truly new heights. When we see what works, we can work to move these things that work in one location for consideration to other parts of the organization.

How to Discover Lessons Learned from Performance Analytics

Performance metrics at the micro and the macro level provide us with sources of understanding. Where our performance is lacking according to our objectives and strategies, we can determine ways to improve this area with focus being brought to the problem. For example, we can bring the organizations talent together to study and devise experiments from which we can learn and improve the performance. We devise plans for experiments using for example that Shewhart or Deming Cycles for the exploration, or some other structured approach.

Sometimes there are limits to being able to improve the performance. There may be one or more constraints within the organization that may limit the ability to internally achieve the objectives. In these cases, we need to think out of the box. When these constraints that are beyond what the organization is capable of adapting, we will need to consider other approaches:

■ outsource

■ joint venture

■ collaboration with schools or other organizations

  • [1] Bothe, Davis, R. (2007). Reducing Process Variation, Using the DOT-STAR Problem SolvingStrategy. ICedarburg, Wisconsin: Landmark Publishing Co. page 490
< Prev   CONTENTS   Source   Next >