Introduction

Why Should We Spend Time on Knowledge Domain Analysis?

As an organization matures in its use of KCS, an important function emerges: Knowledge Domain Analysis (KDA). KDA is part of the Evolve Loop in the KCS double loop model.  It is the reflective half of the model and seeks to understand what we can learn from a collection of knowledge articles - what we call a domain. The KDA activities are critical in maximizing and sustaining the benefits of KCS in two ways.

1. Organizational learning: identifying high impact improvements in processes, policies, and offerings (product and service functionality and usability). 
2. Continuous improvement of the KCS Practices: the content standard, the workflow model, and self-service effectiveness. 

Knowledge Domain Analysis uncovers evidence of the benefits being generated by KCS.  Sharing this evidence with leadership and knowledge workers sustains interest in the KCS activities.

The impact of the Knowledge Domain Analysis is quite broad and includes topics such as: the quality of the articles, the effectiveness of the workflow and problem-solving process, and, perhaps most importantly, the patterns of reuse of the articles.  The patterns and trends of article reuse enable us to identify pervasive issues in our offerings (products and services), processes, or policies. These pervasive issues are candidates for root cause analysis and corrective actions to remove their causes from the environment. 

The initial focus of the KDA activities is on the knowledge base. This analysis assures that issues are resolved effectively and efficiently. However, over time the scope of the analysis expands to include self-service, online communities and social network activity, and content that is related to the domain. This expansion of scope gives us a more complete view of the requestors' experience. 

The success of the Knowledge Domain Analysis function is measured through improvements in findability, self-service use, and self-service success rates, as well as improvements in the customer experience as a result of corrective actions taken to eliminate the cause of pervasive issues. 

Knowledge Domain Analysis outputs include the identification of :

• Improvements to the content standard and process integration (workflow)
• Findability issues, in which knowledge exists but is not being found: search performance and optimization
• Content gaps: knowledge requestors are looking for that does not exist
• Content overlaps: consolidating duplicate articles by identifying the best or preferred resolution among many proposed resolutions  
• Problem-solving of new issues: improvements in how we diagnose and resolve new issues
• Pervasive issues: facilitating root cause analysis and working with the functional owners on corrective actions to eliminate pervasive and high impact issues
• Knowledge base value: article reuse rates, self-service success rates, contributions in reducing time to resolve new issues, and elimination of pervasive issues. 
• Archiving strategy: develop the criteria for when to update an article's confidence state to archived (criteria is often unique to the domain)

When Do We Start Knowledge Domain Analysis?

Early in the KCS journey, as the knowledge workers are learning to capture and reuse their experience in the knowledge base, there are not enough articles or article reuse to produce meaningful patterns.  When do the patterns become interesting? As a general rule of thumb, when the reuse rate of articles becomes greater than the create rate, it is time to start the knowledge domain analysis. This may vary by domain. 

Most organizations have multiple knowledge domains. Knowledge domains are virtual collections of KCS articles that are related to a common topic, function, process, technology, or product family. Knowledge domains are not precise or absolute in their boundaries; they often overlap. A knowledge domain is the collection of content that makes sense to include for pattern recognition and cluster analysis. Therefore, the purpose or intent of the analysis defines the collection of articles that are relevant.

Who Does Knowledge Domain Analysis?

For each domain, one or more Subject Matter Experts (SMEs) emerge as the Knowledge Domain Experts (KDEs) who do or facilitate the Knowledge Domain Analysis activities. KDEs have enthusiasm for and curiosity about the domain topic or function.  The KDEs look after the health of a collection of content or a domain of knowledge, usually a subset of the knowledge base that aligns with their expertise. KDEs must have both a deep subject matter expertise in the domain as well as a profound understanding of KCS. They are knowledge workers who continue to have other functional responsibilities: the KDE is not a full-time role. KDEs are the people who are naturally attracted to using data analytics to figure out what we can learn from a collection of knowledge.

Based on the analysis, KDEs work with the Coaches and the KCS Council to improve the content standard and the KCS workflow.  The KDEs must also be able to articulate the value of the knowledge base and work to provide knowledge workers and management with visibility to the impact of people's contribution to the knowledge base. KDEs also work with the owners of the organization's offerings (products and services), processes, customer success managers, self-service design, and policies who can take corrective actions to eliminate high impact or pervasive issues.   

KDEs must be capable of establishing influential relationships with the various functions that need to take corrective actions. The goal is to provide the functional owner with quantifiable, actionable information that is based on the users' experience. Because of the required cross-functional collaboration, Knowledge Domain Analysis is most effective with cross-organizational participation.

In addition to coordinating the efforts of other SMEs and functional owners, it is extremely helpful for the KDEs to have access to Data Scientists who can help leverage emerging digital automation techniques (sophisticated text analytics, machine learning, and/or artificial intelligence).  The use of digital automation can greatly expedite and enhance Knowledge Domain Analysis capabilities.    

How Do We Do Knowledge Domain Analysis?

The KDE is the driving force in realizing and sustaining the benefits of KCS. The scope of activities the KDE may engage in is very diverse. It includes a mix of doing, facilitating, and influencing.  There are numerous techniques used in Knowledge Domain Analysis including (but not limited to):

• Pareto analysis (frequently used articles)
• Cluster and pattern recognition
• New vs Known Analysis
• Search engine tuning (internal), search engine optimization (SEO, external) 
• Self-service analytics
• Feedback management and analysis on surveys and article feedback
• Root cause analysis  

This guide describes these techniques. While we have great expectations about the value of using emerging digital automation capabilities to support the analysis, here we have described these techniques as manual processes.

The multi-day, interactive Knowledge Domain Analysis workshop covers key elements of the Evolve Loop practices and identifies opportunities for improvements in: KCS Solve Loop processes, high-value knowledge articles, self-service success, and business improvements (products, services, policies).