Knowing what employees know enables companies to pinpoint training and staffing gaps.
Last year, more US employees quit their jobs than ever before, and 2022 is shaping up to be a year of similar churn. That high turnover is shining a light on many organizations' lack of knowledge-transfer and knowledge-retention capabilities. It also is putting a major strain on learning and talent development staff, who must onboard and upskill replacements for departing employees.
Put simply, here's the current state of work and the job market: People are more separated than ever; are staying at their jobs for less time; and have a dearth of resources to help them learn, perform, and grow (both technically and culturally) within an organization. Knowledge management (KM) can play a major role in counteracting those new challenges as well as building a broader foundation of knowledge sharing, collaboration, and innovation within companies that can serve as a true competitive advantage.
Knowledge management defined for today's realities
KM is fundamentally about harnessing an organization's knowledge to ensure that those who need it can reuse it. When approached at an enterprise level, KM involves the people, process, content, culture, and technology necessary to capture, manage, share, and find information. Let's break that down.
- People: the individuals who hold the knowledge, those who need knowledge to grow and perform, and the flow of knowledge between them
- Process: the policies and procedures, roles, and responsibilities governing the capture, management, and maintenance of knowledge and information
- Content: the complete depth and breadth of an organization's knowledge—the tacit knowledge held within experts' heads as well as the complete range of knowledge, information, and data, with a particular focus on digital content
- Culture: company leadership supporting good KM practices and the overall willingness of individuals within the company to share their knowledge
- Technology: the KM systems themselves; the tools that will enable the capture, management, enhancement, and findability of knowledge; as well as the formation of connections between people and the collaboration and creation of new knowledge and information
In the context of today's realities, KM's new mission is to link all an organization's knowledge—in every form—making it not just findable but understandable and actionable.
Through the lens of learning and talent development, an effective KM program ensures that the complete learning and performance ecosystem of materials is findable, accessible, and actionable in the appropriate form for the right employees. That goes beyond searchability, guiding the governance and tagging of learning content in all its forms so that it is accurate and understandable. Beyond digital learning content, KM can also create networks of learning and expertise using expert finders and communities of practice, forging relationships, and creating a ready flow of knowledge from those who have it to those who need it.
Those networks aren't just about learning; they will create new knowledge through collaboration and discourse from which others may engage and develop. In short, good KM will drive L&D and foster collaboration, culture, and innovation. That means improved quality, speed, and consistency of learning alongside higher employee engagement and satisfaction. In addition, employees who feel supported from the first days of onboarding through their career are more likely to stay and grow within the organization and share their knowledge with others, OC Tanner reports. In short, good KM creates happy employees, and happy employees drive good KM.
Knowledge management applied
Artificial intelligence, machine learning, and natural language processing are all enablers of KM transformations. In particular, a new solution—knowledge graphs—has quickly become a major component of many enterprise KM efforts. Knowledge graphs are information hubs designed to mimic the way people think. They identify people, places, and things and map how they relate to one another.
Knowledge graphs are the technology behind many artificial intelligence solutions such as Siri, Alexa, and Google Home. They are typically populated in an automated fashion using machine learning and natural language processes. The real power of knowledge graphs is that they take information in any form and transform it into a structure that both humans and machines can understand to automate activities that used to be highly manual.
Three of the most impact-making solutions based on knowledge graphs within the L&D space are learning path assembly, expertise maps, and predictive staffing. We've put each of those into production for global clients, so they are not the promises of tomorrow—rather, they are the reality of the industry today.
Learning path assembly and adaptive learning
Learning path assembly, or curriculum assembly, enables organizations to create a personalized path for every employee's knowledge-acquisition journey. Each learner-centric path is based on a wide array of factors, including people's backgrounds, education, past experiences, competencies, job goals, and job descriptions. Knowledge graphs are one tool that enables greater automation in training personalization. They excel at aggregating information from multiple sources, both internal and external, to form a more holistic view of each learner.
The graph can retrieve information about a person from their resume (using the recruiting system or standard text analytics), the HR system, LinkedIn, the learning management system, project tools such as SharePoint online, and a time and billing system showing which projects people are working on. Once it gathers the information, the graph knows everything about each individual, including their:
- Work background (from their resume and LinkedIn)
- Current role in the organization (from the HR system)
- Training history (from the LMS)
- Recent work experience (from the time and billing system and the project management platforms)
The graph stores the information in a series of nodes and edges. A node is a thing (such as a person or a course), while an edge is a description of how it is related. For example, a node could be a person's name, an edge could be "has completed," and the other node could be the name of the course (Jane Doe has completed Course X). The course could then have another node that describes one or more skills or subjects the course addresses (Course X addresses Y skill). Another set of nodes could connect the person to a resume and the resume to the skills or certifications listed on the resume.
The graph grows like a spider web with more and more information connecting the individual to information about them. When an L&D practitioner queries the graph, they can retrieve the skills or topics that someone has learned to reveal what the individual knows and how they know it. One of the most powerful things about these graphs is that it is easy to add new pieces of information no matter what format it takes.
L&D teams can use the fully populated knowledge graph either manually or automatically to customize employee learning paths and training. If the course curriculum has training on subjects or skills that a person already knows, the training manager can remove the requirement based on what the graph recommends. If the graph is automating course assembly, it will assemble course curriculums that exclude courses for skills the individual already knows. Knowledge graphs can support either approach.
Ultimately, each individual learner's starting place is factored into the design of an automated learning path assembly solution. That ensures that the L&D function meets every employee exactly where they need it on their upskilling journeys, safeguarding more advanced learners against boredom and providing those who are new to the material a reasonable place to begin.
People learn at different paces and struggle with different concepts. Adaptive learning creates a necessary level of customization and flexibility by integrating basic checkpoints in the form of quizzes, small practical tests, and self-analysis. The checkpoints are integral to gauging the speed of learning that every person requires or responds to best, as well as indicating whether they need more advanced or remedial content in particular areas. The graph can store results from the quizzes and practical tests, further describing the individual's experience.
When an L&D professional queries the graph about a particular piece of training, the graph will compare the topic or skills the training event addresses with the information of what is known about the individual from the graph and then determine whether the training is necessary. There's no reason to force an employee who is excelling in a particular space to complete modules or review materials they have already mastered. The graph enables real-time decisions as to what learners need versus what they already know and understand.
Expertise maps
Expertise finders and knowledge maps have long been a goal of KM programs, but as organizations move further away from full-time, in-office postures, the need to map a company's people and expertise has increased significantly. Many companies talk about expert finders as simple directories that list an individual's contact information and basic skills. However, the problem with most of those is that they either require an employee to list their skills or require the organization to maintain the data in a searchable way.
Both options are rife with error and inconsistencies, especially for large companies. We are now using a combination of knowledge graphs and search to solve this historically complicated problem.
The same knowledge graph that can support learning path assembly can power the expertise finder. Companies can derive expertise levels based on an individual's prior work experience, training, and the customers or projects that the person has worked on since they joined the company. The knowledge graph mines prior work experience from the skills, certifications, and industry experience on each person's resume along with similar information documented on LinkedIn. The training and certifications that the employee has completed internally, as well as the skills and subjects those learning options address, also are storable in the graph. Finally, the graph can scan collaborative project workspaces (such as SharePoint) for the projects the person worked on and the content they wrote.
The knowledge graph collects in an automated fashion the information about what each person knows—there is no need to ask people to manually add the information to the expertise map. As people complete courses or training, that information transfers to the graph on a nightly basis. When people work on new projects, the project along with any information about the topic, customer, or industry the project supports likewise transfers to the graph.
In addition, the graph can scan the content in the project on a regular basis to identify topics the individual may have written about as new areas of expertise. The graph's recommendation algorithm can weigh the topics and subjects from the graph based on the source of the topic to provide an understanding of the level of expertise in any particular area. Experience on a project or formal training would suggest one level of expertise, while content in a document the individual authored would suggest potentially greater knowledge of the subject.
The result is an automated and highly detailed understanding of the skills and topics in which each employee has expertise. The graph can then power search to help employees find the experts across a wide range of topics. The company can also use it to identify the skills of employees who are leaving so that it can plan how to replace their skills or find and train the right next hire.
Predictive staffing
Leading service organizations are now pairing their understanding of company expertise with their employee demographics, sales pipeline, and project management systems. Also known as enterprise 360 in KM terms, this approach enables predictive modeling to identify potential gaps in staffing before they happen.
A predictive staffing solution integrates the customer relationship management system with a skills taxonomy and the knowledge graph to provide proactive recommendations on staffing needs. The predictive staffing solution evaluates current deals of a particular size and average close rate and close time, all from the CRM. It also reviews the skills and competencies required to deliver a similar project to predict upcoming staffing requirements and skills needs.
That information put against the knowledge graph reveals current employee skills and capabilities as well as demographics regarding retirements and other staff attrition. The integrated solution can flag potential skills gaps before they occur, enabling the company to proactively make decisions around upskilling staff or hiring new employees in advance of a staffing shortage.
Reimagined knowledge management practices
Each of the noted solutions is a practical, real-world example of how KM and knowledge graphs are changing how employers operate, making a meaningful impact on employees and proving fundamental for the organizations that integrate KM practices. There is a measurable return on investment to be had here, a value that can save companies time and money while enabling employees and L&D professionals to feel the benefits of KM like never before.
Most employers will have work to do in improving the foundations of their KM—including their knowledge-sharing culture, content quality and structure, process redesign and governance, and core systems—to begin reaping the advanced benefits discussed here. But for companies ready to seize the benefits of the KM rebirth, the time to begin is now.
