Commercial Cartography – Vol. VIII: Cause & Effect Diagrams

     A cause & effect diagram is best conceptualized as a specialized application and extension of an affinity diagram.  Both types of diagram can be used for proactive (e.g. development planning) or reactive (e.g. problem-solving) purposes.  Both use brainstorming techniques to collect information that is sorted into related groups.  Where the two diverge is in the nature of relationships represented.
     An affinity diagram may present several types of relationships among pieces of information collected.  A cause & effect diagram, in contrast, is dedicated to a single relationship and its “direction,” namely, what is cause and what is effect.

     The diagrams presented herein may also be known by other names.  Some use a slight variation, “cause-effect diagram,” or “C-E diagram” for short.  Many know them as “fishbone diagrams” because a common presentation format resembles a fish skeleton.  Others refer to “Ishikawa diagrams” in recognition of Kaoru Ishikawa, the technique’s creator or popularizer, depending on the source cited.  In this discussion, “cause & effect diagram” and “C-E diagram” will be used interchangeably to refer to the output of the exercise described, C-E analysis.  Alternative formats, shown later, demonstrate the advantage of using a universal name for the diagrams.  This allows focus to remain on the content and purpose of the diagram without needless distraction by variations in the display format.
     Numerous variations in the content – particularly the group names or headers – are also possible.  To develop a foundational understanding of the process, a “traditional” C-E diagram development is the focus of the following presentation.  Information on common variations is included as unintrusively as possible.

Go Fish!
Manufacturing
     Cause & effect diagrams originated in manufacturing problem-solving, or troubleshooting; use in this context remains its most common.  The “traditional” set of cause categories – the affinity diagram headers – is known as “the 6Ms,” an alliterative mnemonic.  This set includes:

• Man – or huMan, huwoMan…  Any huperson, really.  By definition, “man” refers to any human being or person.  A jocular jab at modern society’s abuse of the English language should cause no shock, nor concern; feel free to ignore it!
• Machine – or equipMent, broadly defined.  Any mechanical or electronic device, tool, vehicle, etc. employed by a person, or operating autonomously, is included in the machine category.
• Material – or Matter.  Any substance used in the creation of the product, whether a major component or a superficial additive, belongs in the material category.  For example, mold release and die lubricant are not components of the final product, but are critical to their respective manufacturing processes.
• Method.  Any procedure, technique, or ”trick” – whether or not it is delineated in process documentation – is a method for purposes of analysis.
• Measurement.  This category typically references concerns about physical product attributes, such as:
  • Are appropriate properties being measured?
  • Are sample sizes and frequencies sufficient?
  • Are the measurements repeatable and reproducible? (See the “Get Some R&R” installments of “The War on Error” series – “Vol. IV:  Variables” and “Vol. V:  Attributes.”)
    

         Process attributes may also require review.  Examples include temperature, gas flow rate, press force, operator process time, and innumerable others.

• Mother Nature – or environMent.  A useful term for purposes of the mnemonic, it should not be taken too literally.  The environment of concern can be natural or controlled.  It may be on a large scale, such as an entire facility or even the outdoors; it can also be on a small scale, such as an individual room (e.g. laboratory) or inside a heat-treat furnace.

     A generic diagram, using the 6M categories, is shown in Exhibit 1.  The problem statement, or effect to be analyzed, is shown at the end of the center spine.  Branches, or bones, on either side of the spine are labelled with major cause categories.  Lines drawn from these branches are each labelled with a primary cause of the effect identified; secondary causes branch from primary causes, and so on.  Additional levels of “sub-causes” can be added if necessary, using a “5 Why” technique or similar approach to identify the most fundamental cause of the problem under investigation.  A 6M diagram example is shown in Exhibit 2.

     Construction of the diagram progresses following the pattern “[current branch label] is caused by [lower level branch label].”  For example, in Exhibit 2:  “Iron in Product is caused by Materials issues, including raw material issues such as water (H2O) from the city distribution system.”  A statement like this can be used to generate the branches in the diagram.
     To verify proper construction of the diagram, the process can be reversed, reading the diagram according to the following pattern:  “[lower level branch label] causes [next higher level branch label].”  Doing this, the example above becomes “the city’s distribution system contaminates (causes an issue) water (H2O), a raw material for our product (Materials issue) causing Iron in Product.”  The generic pattern is modified to translate abbreviated, truncated, or otherwise simplified notes into an intelligible statement.  The phrasing may be clunky, but it only needs to confirm that causal relationships are properly represented.  Statements like this can be refined for presentation purposes if deemed necessary or useful.
     The troubleshooting and problem-solving example is a “negative effect” diagram, where the goal is to eliminate the negative effect, or problem, identified.  Causes identified can be used to create or improve an FMEA and various controls.
     A “positive effect” C-E diagram can also be constructed by replacing the problem statement with a goal statement.  The diagram is then populated with information that can be used to attain the desired outcome.  Project managers, product development planners, customer service agents, and others can benefit from conducting this type of analysis.
     The preceding presentation describes the cause enumeration type of C-E analysis.  Other types of C-E analysis and diagrams are presented in the following sections.  Later, additional tips for construction and utilization of various diagrams are provided.
 
Service
     Where manufacturing has the 6Ms, service industries have the 4Ss:

• Surroundings – similar to Mother Nature.  Here, the environment is evaluated for its impact on the customer experience, which may differ from a process-performance perspective.  The ease with which a customer can find and engage a service, their comfort level, and aesthetic factors are included in this category.
• Suppliers – similar to Material.  The inputs to a service effect its delivery, including materials, equipment, and other services.
• Systems – similar to Method.  Policies and procedures that define the structure of service and the tools required to deliver consistently are included here.  Examples include order-taking, communication, and verification.
• Skills – similar to Man.  This category identifies the abilities required for successful delivery of a service.  It can be used to develop training plans and evaluation criteria.

     The bones of a service C-E diagram are shown in Exhibit 3 as a negative effect diagram.  A positive effect diagram is created by replacing “The Problem” with a service outcome to be achieved.  A specific characteristic, such as “wait time < 5 minutes,” is usually more productive than a more general, overarching goal, such as “customer delight.”  It may take an entire school of fish to achieve customer delight!

     An example 4S diagram is shown in Exhibit 4.  This diagram has a lot of fishhead (more on this later), but few bones.  Each cause shown should be explored further, expanding the diagram content accordingly.  The diagram’s construction should be double-checked, ensuring that the cause-effect relationships are presented properly.
 
Marketing
     The 4Ps are well-established and widely-known levers of marketing.  The 4Ps are:

• Product – what is on offer.  This could be physical product, a service, or a digital token (e.g. cryptocurrency, NFT).
• Price – cost of the product.  Discounts, rebates, and competitive comparisons are all considered in this category.
• Place – where customers access the product.  Both physical and virtual worlds are considered here.  The position of a photo on a webpage and physical location of a product on a store shelf are analogous; they are chosen for visibility and ease of purchase.
• Promotion – incentives and perception of product.  Advertising, branding, endorsements, sponsorships, and “influencers” are considered in this category.

     Similar to the previous types, a marketing C-E diagram can be used to deconstruct a market failure (negative effect) or to capitalize on a market opportunity (positive effect).  That is, both proactive and reactive uses are again possible.
 
Process
     A process analysis diagram extends the cause enumeration diagram, creating a hybrid.  It combines aspects of a process flow diagram (PFD) and a “traditional” fishbone diagram.  Potential causes of failure, or contributors to success, are identified for each process step.  Traditional cause categories, presented above, can be used, or process-specific categories can be defined.  The process steps themselves can also serve as the major cause categories (discussed further in the next section).  An example process C-E diagram is shown in Exhibit 5.  This structure incorporates an additional ‘level’ of analysis, narrowing the focus of development efforts for each contributor or cause.

Variations
     In addition to cause enumeration and process analysis diagrams, some authors (e.g. Kolarik) cite dispersion analysis as a third distinct diagramming effort.  Dispersion C-E diagrams present causes of product or process variation rather than failure.  However, a cause enumeration diagram with a problem statement referring to excessive or undesired variation achieves, essentially, the same result.  For this reason, dispersion analysis is undifferentiated from cause enumeration in this and many other discussions of C-E diagrams.
 
     The major cause categories presented were described as “traditional” for lack of a better descriptor.  Like other tools, C-E diagrams have evolved and mutated with time and experience.  The 6Ms, commonly associated with manufacturing, have been expanded to include such items as the following:

• Maintenance – often included in the Machine category; a large number of bones may justify creating a separate category for analysis.
• Management – often implied in Method or other cause category or implicated by cause descriptions.  A distinct category may be useful for presentation purposes or assignment of tasks.
• Money – implied everywhere, all the time.  With enough money, virtually any problem can be solved!  On a more serious note, a struggling business or startup may find it useful to separate this category to address cash flow issues, availability of short-term financing, investment opportunities, tax burdens, etc.

 
     Some advocate for adding “safety” to the 4Ss of service C-E analysis.  In addition to the dangers of grouping safety with other concepts or pursuits discussed in “Safety. And 5S.,” there is another flaw in this logic.  “Safety” is typically considered as an outcome or goal, or the lack thereof as a problem to be solved.  Presenting safety as a category of causes distorts perceptions of this most-important topic, leading to poor decision-making.
 
     The 4Ps of marketing have also been expanded for C-E analysis.  Marketing C-E diagrams often include the following additional categories:

• People – or Personnel.  All hupersons still qualify.  Sales staff, “influencers,” trade show booth attendants, or anyone else associated with a business, brand, or product are accounted for here.
• Process – standard product introduction, or “rollout,” campaign templates, etc. that support consistent presentation to customers.
• Packaging – the first interaction a customer has with a product involves its packaging.  Its influence on customer perceptions of quality, capability, ease of use, and so on should be evaluated.  Practical matters, such as product protection, branding, differentiation, etc. must be also be considered.
• Policies – a broad category that includes use of distribution channels, perks for preferred customers, and so on.

 
     Some sources reference the CEDAC – Cause and Effect Diagram with the Addition of Cards.  This “type” of diagram differs only in its method of construction; the content and use remains the same as previously discussed.  C-E diagrams are often constructed by a small, dedicated team in a brainstorming session; this has become such a common assumption that doing otherwise prompted coining a new name.  Rather than relying on a small team in isolation, a CEDAC is constructed in public view and by the entire organization.  Any member of the organization can provide input to the C-E analysis by adding their own “cards” to the diagram.
     Though it is not really a different “type” of diagram, the CEDAC provides significant advantages, such as:

• It utilizes the experience and insight of the entire organization to draw a more-detailed picture of the situation and, thus, make better decisions.  Employee engagement and retention also tend to be higher in an inclusive environment.
• It fosters transparency; problems and decisions are visible to all members of the organization and discussed openly.
• Products, processes, and the influence each person has on their success are better understood by everyone in the organization.
• It promotes widespread, spontaneous training on the technique, increasing the advantages described above.

     Though we have not yet strayed from the fishbone style, cause & effect diagrams may be presented in other formats.  When a C-E diagram is created to seek the root cause of a problem, it may be drawn within a graphic of a tree, as shown in Exhibit 6.  The advantage of this format is that it constantly reinforces the idea of a “root” in participants’ minds; construction is carried out in much the same way as in the fishbone format.  Even in the fishbone format, sub-causes are often referred to as “branches,” reinforcing the idea that these formats are interchangeable.

     A table format C-E diagram is less visually striking, but equally effective once users are familiar with its layout.  It is often generated in a spreadsheet, as no graphics are required; the widely-available tool provides a simple mechanism for storing and sharing the results of a C-E analysis.  A table format is a less-literal version of the tree diagram, as shown in Exhibit 7.

Tips for C-E Diagram Construction
     Realistically, there are few tips needed to begin C-E analysis.  Members of a team assembled to conduct an analysis should have relevant experience; this is not always the same as direct experience.  For example, experience with a process similar, though not identical, to one under scrutiny can be valuable.  Likewise, a team member with experience creating C-E diagrams in manufacturing can help a product development team learn the technique.  Experience is often transferable; teams should not be too quick to exclude those with differing backgrounds.
     If the CEDAC approach is chosen, the entire organization’s experience is applied to the effort.  If it exists in the organization, and the organization has a healthy culture, the necessary information will be included in the diagram.
 
     Regarding the content of a diagram, the cause categories presented are merely suggestions.  Less-experienced individuals or teams often benefit from beginning their analyses with “traditional” categories.  As experience with the technique and the subject of analysis is gained, use of alternative major cause categories may be found to be more effective.  An example C-E diagram that uses alternative categories to good effect is shown in Exhibit 8.  Neither the names nor the number of categories is fixed; adjust as needed!

     One reason to modify the cause categories may be the developing diagram itself.  If there is a large number of bones in any one category, consider how that category may be logically subdivided to make the information more manageable.  If there are very few, determine if it is truly relevant to the objective.  If it is, consider combining it with another category.  A new category name may be needed to accurately represent the combined content.
 
     Many sources of instruction define the format of C-E diagrams much too rigidly.  For example, placement of the problem statement or goal on the right-hand side of the diagram is presented as if doing otherwise nullifies the entire effort.  It does not!
     Sure, we English-speakers read from left to right, but we also write from left to right, which can make whiteboard construction of a diagram awkward if the “rule” is enforced.  Being left-handed may exacerbate this further.  Creating a spreadsheet table in this manner would also be highly inefficient and tedious.  Likewise, the angle of cause branches from the spine, parallelism of fish bones, placement of causes at the ends of bones, and other “rules” are more flexible than they are typically presented to be.
Constructing a C-E diagram can be a fast-paced and messy exercise.  Once complete, a “clean” copy of the diagram is often created for future use.  When this is done, following the conventions (or rigid “rules”) may be advantageous and is often recommended.  Until such time, scribble the information in any manner that prevents it from being lost in the chaos of the exercise.

     Conducting C-E analysis and constructing a diagram follows a process that overlaps with that of creating an affinity diagram.  Brainstorming, sorting, and labelling can be performed in the same manner.  Final arrangement in the diagram results from defining the cause and effect relationships of the captured ideas and further probing for sub-causes.
     When performed for product planning, or other proactive purpose, execution is usually very linear in nature; each step is completed before beginning the next.  Each major cause is thoroughly investigated in succession, identifying data, tests, etc. that will be needed to ensure a successful design.  This information can be used in development plans, DFMEA, and other design documents.
     When performed to aid problem-solving, particularly in urgent situations, tasks may be done concurrently.  The most likely causes are selected by experience and currently-available data that can be immediately and quickly analyzed.  Verification testing and data requirements are identified to confirm that the problem has been eliminated.  Information gathered here can be incorporated in a PFMEA, Control Plan, or other process documentation.

     The graphical formats presented provide visual aids to understanding relationships among the information contained in a diagram.  However, the graphics can become a distraction or a gimmick; instead of aiding comprehension, it begins to interfere with it.  Exhibit 4 was chosen to demonstrate the potential for graphical presentation to run amok.  While the enormous fishhead may attract the attention of an elementary school class, the information that a management team needs to make effective decisions is virtually illegible.  Graphical flourishes are most useful for introductions to the technique and terminology; as familiarity grows, the graphics should fade into the background or be eliminated.

     These tips can be summarized as follows:  use the resources available, in the manner that best suits the organization and the situation, regardless of what “the rules” require.  “The Third Degree” is here to present alternatives and help practitioners understand them; it is not here to impose arbitrary restrictions that provide no meaningful safeguards or other advantage.
 
Summary and Conclusion
     The C-E analysis process in brief:

• Form a team or choose to leave open to everyone.
• Define the problem or goal.
• Choose the type and style of diagram to be created.
• Brainstorm causes or contributors.
• Sort causes or contributors into major categories (cause enumeration) or process steps (process analysis).
• Probe each cause for sub-causes (i.e. root cause).
• Double-check that cause-effect relationships are properly represented.
• Finalize and “clean up” the diagram.
• Utilize the information to solve the target problem or achieve the desired outcome.
• Transfer information to appropriate documents (e.g. FMEA).
• Review the process followed to learn, improve, and teach others.

     Like other tools presented in “The Third Degree,” the cause & effect diagram appears simple, but has a lot of hidden value.  Readers are encouraged to explore this and other “simple” tools to discover new applications or modes of thinking about their objectives.  Creative adaptation of existing tools and techniques is far better than reinventing the wheel.
 
     For additional guidance or assistance with Operations challenges, feel free to leave a comment, contact JayWink Solutions, or schedule an appointment.

     For a directory of “Commercial Cartography” volumes on “The Third Degree,” see Vol. I: An Introduction to Business Mapping (25Sep2019).
 
References
[Link] “What is a Cause and Effect Diagram and How to Make One.”  Smartdraw.com
[Link] “What is a Fishbone Diagram?”  American Society for Quality, ASQ.org
[Link] “9+ Fishbone Diagram Teemplates - PDF, DOC.”  template.net
[Link] Creating Quality.  William J. Kolarik; McGraw-Hill, Inc., 1995. 
[Link] “Guide To Root Cause Analysis - Steps, Techniques & Examples.”  Software Testing Help.
[Link] “The Ultimate Guide to Cause and Effect Diagrams.”  Juran.
[Link] The Six Sigma Memory Jogger II.  Michael Brassard, Lynda Finn, Dana Ginn, Diane Ritter; GOAL/QPC, 2002.
 
Jody W. Phelps, MSc, PMP®, MBA
Principal Consultant
JayWink Solutions, LLC
jody@jaywink.com