Knowledge management is how an organization creates, approves, organizes, shares, and maintains the information people need to do their work. The practical test is simple: can someone find a current, trustworthy answer and know what to do when it does not cover their situation?
Let’s explore a familiar scenario. An employee has a question. It’s not an unusual question, and the situation is not especially complicated. They just need accurate information to complete a task, make a decision, or help a customer.
They begin with the usual sources. First, they search through the notes they have accumulated over time: the handwritten reminders, saved emails, bookmarked pages, and documents stored on their computer.
Nothing.
Next, they check the usually reliable Frequently Asked Questions section on your SharePoint site.
Still nothing.
They ask a coworker. That person is unsure, so the employee asks another coworker. Then another. Each person offers a slightly different interpretation, but no one can point to a reliable source that confirms the answer.
Eventually, after exhausting the usual options, the employee takes what can sometimes feel like the “nuclear option”: they ask a manager.
Unfortunately, the manager doesn’t know either.
Wanting to be helpful, and feeling pressure to provide an immediate answer, the manager makes an educated guess. It sounds reasonable, but it is based on memory or interpretation rather than a current authoritative source. The desire to help creates the risk of being “fast and wrong.”
Regardless, the employee finally has an answer, but they don’t like it.
So, they ask another manager. Then another.
Eventually, someone provides an answer that feels more practical, more favorable, or more consistent with what the employee hoped to hear. The search ends with a palatable answer, without establishing whether it is the correct one.
This is “answer shopping.”
It may look like an employee behavior problem or a management failure. In this scenario, there is also a deeper issue.
The information failed them.
The employee and managers are responding to a broken information system. Blaming the employee for asking again, or the managers for inconsistent answers, leaves that system unchanged.
The employee needed information and used every channel available to find it. The managers were placed in a position where they were expected to provide certainty without access to a trustworthy source.
Clear, current guidance could have resolved the routine question much earlier. A shared source would also have helped managers distinguish an established answer from a situation that genuinely required judgment.
Instead, the organization created an environment in which the answer depended on whom the employee asked.
That’s a knowledge management problem.
There is an important distinction here. Asking for clarification, challenging an error, reporting a concern, or escalating an exception is not the same as seeking a preferred answer after a clear decision. Keep those legitimate routes open. If guidance is clear and applicable but someone repeatedly bypasses it, address that behavior separately. A knowledge system should support sound judgment, not make an incorrect answer impossible to challenge.
What is knowledge management?
Knowledge management is ongoing work. Publishing a document is only the beginning; someone has to keep it accurate, usable, and connected to how decisions are made.
Some knowledge is already written down. Some lives in the experience of people who know which detail matters or when the usual procedure will fail. Capture that experience through worked examples, observation, and conversations, then have the responsible owner validate the guidance before treating it as official.
Effective knowledge management answers practical questions such as:
- Where should employees look for official information?
- Which version of the policy is current?
- Who is responsible for approving and updating it?
- How frequently should the information be reviewed?
- What happens when an employee identifies an error or gap?
- How can leaders confirm that outdated guidance is no longer being used?
- Which questions are employees repeatedly asking, and what critical knowledge do those questions reveal should be documented?
Without clear answers to these questions, information becomes fragmented. These same foundations help new employees learn how to find guidance instead of inheriting another collection of personal notes.
What knowledge chaos looks like
In many organizations, important knowledge is scattered across a broad range of sources and formats, including a mixture of:
- Handwritten notes
- Personal hard drives
- Shared folders
- Saved emails
- Employee handbooks
- Training presentations
- Frequently asked questions
- Intranet pages
- SharePoint sites
- Messaging platforms
- Informal instructions passed from one employee to another
- Documents with names such as “Final,” “Final Updated,” and “Final Version 3.1.2.5”
Some of this information may have been accurate when it was created. The problem is that policies, systems, products, regulations, and customer expectations change.
A handwritten note does not update itself when a policy changes. A downloaded document does not alert its owner when a new version is published. An old training presentation may continue circulating long after the underlying process has been revised.
Over time, the organization accumulates competing versions of the truth. Start an audit with one recurring question and trace every place its answer appears. The differences will show you where ownership or updates have broken down.
One manager relies on the employee handbook. Another remembers an email sent six months ago. A third consults an outdated file on a shared drive. A fourth follows the process that worked at their previous employer.
Each person may sincerely believe they are providing the correct answer.
The customer or employee receiving that answer has no reliable way to know.
The importance of knowledge management for AI
Adding an AI assistant does not settle which version of a policy is authoritative. If an assistant retrieves answers from internal documents, the organization still needs to decide which sources it may use, who can access them, and when outdated material must be removed from retrieval.
Retrieving a document at the time of a question is different from retraining a model. Do not assume an uploaded correction has updated every system or erased an older answer. Verify how the particular tool uses and refreshes its sources.
A current knowledge base does not guarantee a correct generated answer. NIST’s Generative AI Profile identifies the risk of confidently presented false content, including misleading citations. Require source links where possible, and check that the cited passage actually supports the answer.
Before deployment, test real questions, conflicting documents, missing answers, and questions the user should not be permitted to access. Keep source permissions intact. Define when the assistant should decline to answer or send the question to a responsible person, especially for consequential decisions.
Continue reviewing errors after launch. The goal is useful assistance with accountable human ownership, not another confident voice in the answer-shopping chain.
The business consequences are significant
Poor knowledge management may look like a documentation issue. Its effects show up in the work.
Inefficient processes
Employees spend valuable time searching for information, comparing conflicting sources, and asking coworkers for help.
Track the time spent searching and the number of people pulled into recurring questions. That gives you a business-specific baseline instead of assuming a generic productivity saving.
Inconsistent decisions
When managers use personal judgment to fill gaps in organizational knowledge, similar situations can produce different outcomes.
Employees notice these inconsistencies. Once they learn that answers vary by manager, they may continue asking until they receive the result they prefer.
Poor customer experiences
Customers expect accurate and consistent service.
If the answer changes depending on which employee handles the request, trust begins to erode. Customers may need to repeat themselves, reopen cases, escalate concerns, or challenge information they have already received.
Increased organizational risk
Outdated or inaccurate guidance can lead to incorrect commitments, inconsistent handling of concerns, and avoidable business risk. For guidance involving legal or regulatory obligations, involve the appropriate qualified reviewer rather than treating an internal knowledge article as legal clearance.
When an organization cannot identify the authoritative version of a policy or process, it also becomes difficult to demonstrate that employees were given correct guidance.
Frustrated managers
Managers should spend their time coaching employees, improving performance, solving meaningful problems, and supporting business priorities, which is easier in a culture of accountability and ownership.
Instead, they become human search engines for information that the organization should already make accessible.
When managers cannot find reliable answers, they need a named escalation route and a way to tell the employee when to expect a response. Guessing should not be the default service standard.
Frustrated employees and fragile knowledge
Employees want to perform well. Repeatedly encountering unclear instructions and contradictory answers makes that harder.
Repeatedly encountering contradictory answers can erode trust in official resources. Employees create personal workarounds and depend more heavily on informal networks. When a knowledgeable employee changes roles or leaves, undocumented decisions and practical know-how can leave with them. That risk is about how knowledge is maintained, not an employee’s age.
The cycle reinforces itself: the less people trust the official resource, the more unofficial guidance they create. Fixing the resource and showing that reported gaps get resolved can help rebuild that trust.
Creating order does not require an immediate transformation
Organizations often assume that improving knowledge management requires a massive technology project. Technology can help, but a new platform alone will not solve the problem.
If unclear, duplicated, and outdated information is moved into a new system without governance, the organization has simply relocated the chaos.
The work should begin with a few practical foundations.
1. Establish a trusted source for each kind of information
Employees should know exactly where to find official information.
That does not necessarily mean every piece of organizational content must live in one enormous repository. It means there should be a clearly defined source for each type of authoritative information.
When multiple systems are necessary, employees should still have a simple and consistent way to locate the correct resource.
2. Assign an owner to every critical resource
Every policy, process, procedure, knowledge article, and internal documentation asset should have a named owner.
Ownership means more than uploading a document. The owner is accountable for resolving questions, coordinating approval, and updating the content when circumstances change. Make the owner, approver, effective date, and next review date visible on critical resources.
Keep a simple ownership register: resource or topic; authoritative link; owner and backup; approver; intended audience; effective version; next review date; and escalation route. The register should point to the approved content, not become another competing copy of it.
3. Review on a schedule and when something changes
Critical information should have a defined review date.
Set review intervals according to the topic’s risk and rate of change. Also define change triggers, such as a revised product, system, policy, or applicable requirement. A future review date is not a reason to leave known errors in place.
A review schedule turns knowledge maintenance into a routine business process rather than an emergency cleanup project.
4. Remove duplicates from active use and archive old versions
Publishing a new policy is only part of the work. Identify competing copies, label superseded versions clearly, and update the links and training materials that still point to them.
Archive records according to applicable retention requirements and preservation obligations; do not simply delete every old version. Make the current guidance easy to distinguish from historical records, and tell affected employees what changed.
5. Make information easy to find and understand
A technically accurate policy has limited value if employees cannot find or interpret it.
Content should use clear language, intuitive labels, and terms employees are likely to search for. Important information should be organized around the questions employees are trying to answer and the tasks they are trying to complete. Permissions management should control who can view or edit sensitive internal documentation while keeping appropriate knowledge broadly accessible.
The goal is not simply to store information. The goal is to make that information usable.
Test findability with the people who use the guidance. Give them a real question without telling them where to click. Watch the terms they search, the version they choose, and whether they can explain the next step. Revise the resource around what the test reveals.
6. Create a feedback process
Employees are often the first people to discover that instructions are incomplete, outdated, or confusing.
Give employees a simple way to flag an error, request missing guidance, or challenge an answer that does not fit. Name the person who triages those reports and set response expectations based on urgency. Send the resolution back to the person who raised it, then update the approved source when appropriate.
A resolved ticket can reveal useful new guidance. Validate it before publishing, and remove customer or employee details that do not belong in a broadly accessible resource.
7. Pay attention to repeated questions
Every repeated question is useful data.
If employees keep asking about the same issue, the relevant guidance may be missing, difficult to locate, poorly written, or inconsistent with actual practice. Check those possibilities before assigning more training.
Track failed searches, recurring escalations, confirmed conflicting answers, and critical resources past their review date. Sample whether people can find the current answer and use it correctly. Fewer questions alone is not proof of improvement; employees may simply have stopped asking.
8. Reinforce the system through leadership
Employees will not trust an official resource if managers continue offering answers based solely on memory or personal preference.
Leaders should model the expected behavior by consulting the approved source, linking employees to it, and reporting gaps. When an exception is justified, record who approved it, why it applies, and whether the underlying guidance needs to change. An exception should not silently become a competing rule.
The objective is not for managers to memorize every answer. It is for managers and employees to know how to find the right answer.
The right answer should not depend on whom you happen to ask. When judgment is required, employees should know who has the authority to make that decision and how its reasoning will be recorded.
Answer shopping is a symptom.
Answer shopping can signal that people cannot find or trust the information, or that they are seeking a different outcome. Determine which problem you have. A clearer resource will not replace accountability for intentionally bypassing a valid decision, and an accountability conversation will not repair missing guidance.
Start with one high-friction topic. Identify the authoritative answer and its owner, resolve conflicting versions, test whether employees can find it, and review the questions that remain.
The goal is not to eliminate every question. Questions are essential to learning and good judgment. The goal is to stop making employees navigate conflicting information to perform routine work, while preserving a clear route for concerns and exceptions.
Connect with Peoplyst to begin
At Peoplyst, we help organizations examine the systems and decisions that make work harder than it needs to be. If fragmented knowledge or inconsistent guidance is getting in the way, schedule a conversation with Tearle about clearer ownership, usable guidance, and practical next steps.
Connect with Tearle Johnson on LinkedIn for more on operational effectiveness and the systems that help people do their best work.
