Commercial Cartography – Vol. VII: Affinity Diagrams

     An affinity diagram may stretch the definition of “map,” but perhaps not as much as it first appears.  Affinity diagrams map regions of thought, or attention, within a world of unorganized data and information.
     Mapping regions of thought in an affinity diagram can aid various types of projects, including product or service development, process development or troubleshooting, logistics, marketing, and safety, health, and environmental sustainability initiatives.  In short, nearly any problem or opportunity an organization faces can benefit from the use of this simple tool.
     Numerous resources providing descriptions of affinity diagrams are available.  It is the aim of “The Third Degree” to provide a more helpful resource than these often bland or confusing articles by adding nuance, insight, and tips for effective use of the tools discussed in these pages.  In this tradition, the discussion of affinity diagrams that follows presents alternative approaches with the aim of maximizing the tool’s utility to individuals and organizations.

     It is a simple process to create an affinity diagram; much of it is familiar to readers that have used other tools or “maps.”  Similar to a service blueprint, a large table, whiteboard, or wall should be used to allow several people to engage in the physical creation process simultaneously.  Use of sticky notes or index cards and masking tape allows rapid rearrangement of information.  Use of markers and large print allows participants to read cards from a distance of several feet.
     Creating an affinity diagram begins with a brainstorming session, or similar information-collection exercise, conducted with a cross-functional team of 5 – 7 members.  If more are needed to capture all relevant viewpoints, consider conducting multiple sessions, consolidating information in a later step.
     A facilitator is assigned to commence and monitor the proceedings, keeping the group on topic and on schedule.  To commence a session, the facilitator poses a concise question to the team, displaying it prominently in the group workspace.  The question can be related to a problem to be solved, a new opportunity, or any topic for which group consideration could be helpful.
     Each member of the group records responses to the question on sticky notes or index cards (“cards”), adding each to a central collection.  Typical brainstorming rules should be followed.  Particularly important is to refrain from discussion and debate; free-flowing, uninhibited ideas are needed.  If the team or facilitator prefers, all contributions could be recorded on a whiteboard, flipchart, etc. and transferred to cards in the next phase.  This approach may be advantageous if there is concern that a slower process may cause some ideas to be lost.
     When all of the group’s responses have been recorded, the team then begins to organize the cards by clustering similar or related ideas.  That is, ideas are grouped according to their affinity.  Participants may view these relationships differently; a card may be moved several times before there is agreement on the most appropriate placement.  If no consensus can be reached in the initial sorting process, a card may be duplicated; the idea is then included in multiple groups for further consideration.
     During the reorganization, or after it is complete, a label that describes each group becomes evident.  These labels are placed on “header cards” to identify each affinity group.  Any member of the group can create a header card or edit one created by another member.  Every card need not be grouped with others; standalone ideas are normal and permissible.
     Affinity groups can be joined under “superheaders” or further divided under “subheaders.”  The level of “filtration” most appropriate depends on the size of the groups and the way in which collected information will be used.  No rule can be defined; the only guideline is to do what works best for the team engaged and the objectives pursued.
     Many tutorials on affinity diagrams instruct the team to sort the cards without speaking.  There are several issues with an enforced silence approach, however, including:

• It is unnatural among a group of familiar people.
• It delays the discovery of team cohesion issues by facilitating passive-aggressive behavior.
• It treats professionals like children, turning the facilitator into an elementary school librarian.  Constant “shushing” is more offputting than the dialogue it seeks to suppress.
• It precludes clarification of ideas for efficient grouping.
• It delays decisions to duplicate cards, slowing the sorting step.

It is preferable for the facilitator to advise the team to minimize the “noise,” limiting conversations to clarifying questions or other brief exchanges that accelerate development of the diagram.  Detailed discussions of the merits of ideas, priorities, etc. should be quashed until the sorting step is complete.  A pictorial summary of the process thus far is presented in Exhibit 1.

     Presenting a wide range of ideas in this way assures every member of the cross-functional team is cognizant of the entire scope of a situation.  Too often, employees in siloed departments are unaware of the efforts taking place outside their area of responsibility, leading to “turf wars.” Departments vie for resources without understanding the implications for the organization as a whole.  An affinity diagram can provide the “big picture” understanding needed to encourage teamwork that leads to improved decision-making and resource allocations.
 
     After the cards have been sorted and labeled, the ideas in each group are subjected to closer scrutiny.  Again, this parallels a traditional brainstorming session, where contributions are evaluated only after collection is complete.  For smaller diagrams, all contributions may be evaluated by the entire team; clusters within large diagrams may be assigned to a subgroup for review.  These reviews may occur concurrently, as a component of the structured exercise, or independently.  Also, an individual may convene his/her functional department to review the cards and refine the information gathered.
     During the review, duplicate ideas are eliminated and those that are very similar may be combined.  The information on a card could also be split if doing so makes it more manageable.  The merit of each idea is thoroughly assessed; unrealistic or infeasible ideas are set aside.  Remaining ideas are prioritized and more fully developed.  Tasks and projects are defined based on the priorities established.  Cards that had been set aside are placed at the bottom of the stack – literally and figuratively – to be reprioritized should changing conditions effect their feasibility.
 
     Like other large-format map creations, it is possible to simply leave a finalized affinity diagram on a conference room wall.  However, subsequent activity spurred by the diagram, as well as the maintenance of organization-wide awareness, is aided by transforming it into an easily-distributed format.  This often means transcribing the information in an electronic document of some kind.  Common programs, such as spreadsheets and presentation-builders, can be used; single-function software or online tools are also available for this purpose.  In the simplest case, a digital photograph of a wall-sized diagram can be shared, provided the cards are legible in it.
     An example of what a transcribed affinity diagram may look like is presented in Exhibit 2.  The color code often recommended is depicted:  yellow for each idea, blue for headers, and pink for superheaders (or pink for headers and blue for subheaders, if this terminology is preferred).  Other colors may also be used to identify additional levels of filtration.  The choice of colors is not critical, but should remain internally consistent to avoid confusion.

     The example in Exhibit 2 is greatly simplified; an affinity diagram can have dozens or hundreds of items in any number of categories and levels.  Nonetheless, it demonstrates the flexibility of the tool.  Headers and subheaders can be organized differently depending on the phrasing of the foundational question, participants’ points of view, and so on.  For example, a railroad museum is grouped under Man-Made/Transportation, suggesting an interest in machinery and construction.  It could also have been grouped under the Historical header to reflect an overriding interest in the impact railroads had on the culture and economy of SC.  The railroad museum could also have been placed in both groups.  The objectives to be pursued ultimately determine what arrangement is most appropriate; the facilitator provides insight, guiding the team to a consensus.
     Another example diagram is shown in Exhibit 3; though no color code is used, the information is easy to read in this format.  The labels chosen for grouping issues related to medication delivery are associated with components of the system.  It would be understandable had this team, in its haste to improve patient care, skipped a step.  The issues identified could have been sorted under headers such as doctor, nurse, pharmacist, and administrator, identifying the party responsible for resolution of each.  However, this approach could stifle creativity and collaboration in the search for solutions; it is best to assign responsibility only after the logical structure of the diagram has been established and agreed upon by team members.

     While a distributed document is helpful, the physical artifact can still be useful.  If the clusters within a diagram are used to make task or project assignments, for example, a stack of cards, with a header card on top to identify its contents, can be placed in the hands of the responsible individual.  For some, a physical artifact provides motivation and inspiration that an electronic document simply cannot match.  Each card also provides a convenient place to jot additional notes, make sketches, etc. for use in the activity it represents.
     A team may want to consider various sorting and labeling schemes before finalizing an affinity diagram.  The caveat regarding premature assignment of responsibility notwithstanding, viewing the collection of information in various configurations can provide further insight into the challenges the team will face when applying a finalized diagram to chosen objectives.
     One method of refining a problem definition is to combine subgroups until a minimum number of superheaders is attained.  If the information can be distilled to two orthogonal sets, the situation is reframed, facilitating decision-making and prioritization.
     To illustrate this example, consider the schematic in Exhibit 4 that represents a generic customer complaint review.  Various types of issues can be identified – durability, documentation, speed, customization, and on and on.  And on.  Thorough consideration of these issues, in this example, reveal that two types of customer are not well-served:  novices and superusers.

     This revelation reveals that a fundamental decision must be made before the information in the diagram can be effectively utilized:  what market(s) will be targeted?  The following choices are available:

• Expand capabilities to satisfy both novices and superusers.
• Develop product/service to cater to superusers.
• Develop product/service to cater to novices.
• Replace current offering with two separate products/services; one for each user type defined.
• Maintain existing product/service offering (the “do nothing” option).
• Withdraw the product/service from the market with no replacement.

Had this bifurcation of issues not been discovered, efforts to address customer issues could be misguided or haphazard.  Resources could be wasted with no discernible improvement in performance.  This example is applicable to product/service development and marketing efforts.
 
     Though development of an affinity diagram has been presented as a mostly linear process, like others discussed in these pages, this is not strictly true.  Idea generation and sorting often overlap; each activity ebbs and flows within the overall effort.  Experienced practitioners are more likely to do this naturally, as relationships between ideas are more readily apparent and cohesive teams have been established.
     It has also been presented as a group exercise; teams of coworkers addressing business issues is the archetypal application.  However, affinity diagrams are equally applicable to individual pursuits.  Even a team of one benefits from organized information, structured review, potential reframing of an issue, and other discoveries that can inspire creative responses to a difficult situation.
     An affinity diagram is a deceptively simple tool.  Despite its apparent lack of complexity, the insights that can be derived from it and the purposes for which it can be utilized are vast and impressive.  Simply put, organized information is easier to process and utilize.  When a more rigorous collection and refinement process is used, it may be referred to as the KJ Method. This name honors its creator, Jiro Kawakita, a Japanese anthropologist, who formalized the method in the 1960s.
 
     For additional guidance or assistance with mapping or other 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] The Six Sigma Memory Jogger II.  Michael Brassard, Lynda Finn, Dana Ginn, Diane Ritter; GOAL/QPC, 2002.
[Link] “What is an Affinity Diagram?”  American Society for Quality.
[Link] “Affinity diagram.”  Productivity-Quality Systems, Inc.
[Link] “Affinity Diagrams: How to Cluster Your Ideas and Reveal Insights.”  Interaction Design Foundation.

Jody W. Phelps, MSc, PMP®, MBA
Principal Consultant
JayWink Solutions, LLC
jody@jaywink.com