7 best practices for knowledge management strategies

1. Start with a business case

Organizations must begin with a business case, according to Lisa Sanders-Nakahara, head of global operations, marketing and people at Calyptia Inc., an observability platform vendor. A business case can help knowledge teams determine the resources required to create or update a successful knowledge management program.

To start, teams can create an objectives list that includes the organization's knowledge management challenges and the opportunities a program can create for employees and the business at large.

Examples of goals that Sanders-Nakahara found helpful include the following:

• Break down silos between departments.
• Streamline how employees can find and use knowledge resources.
• Create opportunities for employees to actively contribute to the knowledge base.
• Speed up onboarding of new employees.

2. Create a knowledge audit

Next, organizations should start a knowledge audit. This is an enterprise-wide survey of knowledge assets -- such as training materials, onboarding reference guides, process documents and other resources employees regularly use to do their jobs.

To begin, knowledge teams can survey employees about the resources they find most beneficial. Employees can also identify areas where they want or need better documentation to improve customer satisfaction.

A knowledge audit can help determine where current knowledge assets live, how employees use them and potential gaps, Sanders-Nakahara said. She also recommended talking to the people that help define the processes and choose tools that support the new knowledge management program for their insights and goals for the program.

Often, a knowledge management program requires a shift in culture. "You want to be sure that you have the organizational maturity and a culture of knowledge engagement to implement the program well," she said. "A well-executed and maintained [knowledge management] program can help trigger imagination and innovation by creating a continuous source of ideas and experiences."

3. Implement a semantic ontology

Many organizations use different terms to describe the same things, which can create challenges for knowledge teams. Brian Platz, CEO and co-founder of Fluree, a graph database company, said he recommends organizations create semantic ontologies with universal standards -- or a business-wide dictionary -- to make information accessible, interoperable and reusable.

A semantic ontology is a dictionary that standardizes language across documentation, data sources and applications. For example, the World Wide Web Consortium standards group created a semantic ontology in its resource description framework (RDF) data schemas. RDF enables Google to pull a restaurant's open hours, type of food and address into a quick summary.

In an organization, ontologies can help employees find information within a knowledge base more easily. For example, if someone wanted to edit a customer profile, that option would appear higher than unrelated documents that simply mention editing and customers.

A semantic ontology could also speed up data collaboration and sharing that would otherwise be slowed due to translation across domains, Platz said. For example, different departments can agree on what terms can describe job skills to aid in recruiting, onboarding and upskilling.

4. Focus on change management

A successful knowledge management program must address ongoing human processes as much as the tools. Strong and ongoing change management focused on changing behaviors can help build a successful knowledge management strategy, according to Kathy Rudy, chief data and analytics officer at technology advisory firm ISG.

"Knowledge management is not something you roll out with a big bang, hold a few training sessions on and call it a day," Rudy said. "The day never ends."

Knowledge management is not something you roll out with a big bang, hold a few training sessions on and call it a day … The day never ends.

Rudy's team found that the tools and processes were the easy part, but getting people to embrace the tools and process takes more time. ISG is two years into rolling out a new knowledge management system and process, yet many areas of the business still lag behind, according to Rudy. For example, many people view learning the new system as a low priority until they can't find something.

5. Assign responsibility

Organizations should set up a dedicated team to continually work with the people that create, refine and retire knowledge. An ownership map that assigns responsibility for different types of knowledge can help, Rudy said.

Knowledge teams should also create a schedule for maintenance. They can keep tabs on the system's progress and call out problems when they arise. These teams can be a little pushy, too, when necessary. For example, if stakeholders don't respond, teams should be able to remove dated information that no one has accessed over a predefined period of time.

6. Update metadata

Metadata tags provide high-level descriptions of data in documents. If knowledge teams continually refine metadata tags as the business changes how it operates -- ideally with an automated retagging process -- they can ensure all knowledge is up to date and accurate, so employees can easily find necessary information.

"We regularly add new solutions and refine our organization, so our data tags need to match the new business taxonomy," Rudy said. Otherwise, finding documents and data becomes impossible.

7. Create a single source of truth

Different lines of business usually generate data and information that live in their own siloed systems. However, as businesses scale and embrace digital transformation, decision-makers across the organization can also benefit from this data.

Organizations should create a single source of truth (SSOT), according to Andrew Filev, CEO and founder of Wrike, a project management tools provider. SSOTs can help employees more easily find information they need and can minimize confusion.

However, knowledge teams may struggle to get buy-in across different departments -- particularly if they are used to storing knowledge across various tools. Filev said he recommends getting the leaders and teams involved in the process from end to end.

"Early engagement can make sure that these decision-makers understand the project, why it is needed, the information that will inform it, and, ultimately, better recognize and appreciate the outcomes for all teams," Filev said.