Key Takeaways
- A knowledge audit is the necessary first step to fixing a broken information ecosystem.
- By systematically inventorying where knowledge lives—across drives, wikis, inboxes, and people's heads—you can see the full picture for the first time.
- Assessing quality means evaluating currency, findability, and actual usage—not just whether a document exists.
- Identifying critical gaps, especially knowledge trapped in individuals, reveals your organization's biggest vulnerabilities.
- The result is a prioritized roadmap for building a knowledge system that actually works.
You can't fix what you can't see.
Before you can improve how your organization manages knowledge—whether building an internal knowledge base or deploying AI—you need to understand what you actually have. Where does information live? What's current and what's stale? What's missing? Where are people struggling to find things?
A knowledge audit answers these questions. It's not glamorous work—it's spreadsheets and interviews and catalog reviews. But it's the foundation for every knowledge management improvement that follows. Without it, you're guessing. And guessing leads to building systems that replicate the same problems you already have.
This article walks you through a practical audit process and includes a template you can use to get started.
The Three Goals of an Audit
Before diving into the steps, it helps to understand what you're actually trying to accomplish. A knowledge audit has three goals:
Goal 1: Visibility
You need a clear picture of where knowledge currently lives—every shared drive, wiki, intranet page, email thread, and person's head. Most organizations are shocked by how fragmented their knowledge landscape is.
Goal 2: Assessment
Not all knowledge is equal. Some documents are current and useful. Others are outdated, duplicated, or contradictory. You need to evaluate what you have, not just catalog it.
Goal 3: Action
The point of the audit isn't a report that sits on a shelf. It's a prioritized list of what to fix, what to create, and what to retire—so you can make meaningful improvements with limited resources.
Step 1: Map the Landscape
Start by identifying every place knowledge lives in your organization. This is broader than most people expect. You're looking for:
- Formal repositories: Shared drives, document management systems, wikis, intranets, knowledge bases
- Semi-formal sources: Slack channels, Teams conversations, email threads, project management tools
- Informal sources: Individual hard drives, personal notebooks, bookmarked links, people's memories
- External sources: Vendor documentation, regulatory references, industry resources your team relies on
For each source, document what type of knowledge it contains, who maintains it, and roughly how much content is there. Don't try to review every document at this stage—you're drawing the map, not exploring every trail.
Most organizations discover that their knowledge is spread across 5 to 15 different systems—many of which leadership didn't even know were being used.
Pay special attention to the informal sources. These are the ones that represent the biggest risk. When knowledge lives in someone's inbox or in their head, it's one departure away from being gone forever. If this sounds familiar, you've likely encountered the hidden cost of relying on individual experts.
Step 2: Assess Quality
Once you know where knowledge lives, you need to evaluate its condition. For each major knowledge source or repository, assess:
- Currency: When was it last updated? Is the information still accurate?
- Completeness: Does it cover the topic adequately, or is it partial?
- Findability: Can someone who needs this information actually find it? Is it organized logically? Is it searchable?
- Usage: Is anyone actually using it? Some organizations maintain elaborate documentation that nobody reads.
- Consistency: Does this source agree with other sources on the same topic, or are there contradictions?
You don't need to assess every single document. Sample each repository. Pull 10 to 20 representative documents and evaluate them. That gives you a reliable picture without turning the audit into a multi-year project.
In many organizations, 60 to 70 percent of documented knowledge is outdated, duplicated, or effectively unfindable. The audit is what makes this visible.
(Industry estimate)Step 3: Find the Gaps
This is where the audit gets really valuable. Gaps are the topics, processes, or knowledge areas where documentation doesn't exist at all—or exists but is so outdated or incomplete that it might as well not.
To find gaps, compare what's documented against what people actually need. Good sources of gap information include:
- New hire questions: What do new employees ask about in their first 90 days? If those questions don't have documented answers, that's a gap.
- Support tickets and help requests: What do people ask IT, HR, or operations about repeatedly?
- Escalations: Which questions can only be answered by specific senior people?
- Process breakdowns: Where do mistakes happen most often? Poor documentation is frequently the root cause.
Interview a cross-section of employees. Ask them: "When was the last time you couldn't find information you needed? What was it? What did you do instead?" The answers will reveal your most critical gaps.
Step 4: Identify People Dependencies
This step is related to gap analysis but deserves its own focus. You're looking for knowledge that exists only inside specific people's heads—the critical distinction between information and knowledge. This expertise would be lost if those individuals left.
For each department or team, ask:
- Who is the go-to person for questions about specific topics?
- What would break or stall if that person were unavailable for two weeks?
- Are there processes that only one person fully understands?
- Is there historical context or decision rationale that only certain people remember?
People dependencies are your highest-risk knowledge gaps. Unlike a missing document, which you can create, knowledge locked in someone's head can disappear permanently and without warning.
Document these dependencies clearly. Name the person, describe the knowledge they hold, assess how critical it is, and note whether any of it has been captured elsewhere. This becomes your urgency list.
Step 5: Prioritize and Plan
By now you have a lot of information. You know where knowledge lives, what condition it's in, where the gaps are, and which people represent single points of failure. The final step is turning all of that into an action plan.
Prioritize using two dimensions:
- Impact: How many people are affected? How critical is this knowledge to operations, compliance, or revenue?
- Risk: How likely is this knowledge to be lost? Is it held by one person nearing retirement? Is it in a system being decommissioned?
High-impact, high-risk items go to the top of the list. These are the things you address first—not because they're easiest, but because they matter most.
For each priority item, define a clear action:
- Create: This knowledge doesn't exist in documented form. Someone needs to write it, record it, or otherwise capture it.
- Update: Documentation exists but is outdated. It needs to be reviewed and refreshed.
- Consolidate: Information exists in multiple places, possibly with contradictions. It needs to be merged into a single authoritative source.
- Retire: This content is no longer relevant and should be archived or removed so it doesn't create confusion.
- Transfer: This knowledge lives in one person's head and needs to be captured before it's at risk.
Knowledge Audit Template
Use the following five-part template to structure your audit. Each part corresponds to a step in the process above.
Part 1: Knowledge Inventory
| Repository / Source | Type | Owner / Maintainer | Content Description | Approx. Volume | Last Updated |
|---|---|---|---|---|---|
| Company Wiki | Formal | IT / All teams | Policies, procedures, how-tos | ~500 pages | Varies |
| Shared Drive (G:) | Formal | Department leads | Templates, reports, project files | ~12,000 files | Varies |
| Slack #general-questions | Semi-formal | None | Ad hoc Q&A, troubleshooting | ~200 msgs/week | Ongoing |
| Individual email / notes | Informal | Individual employees | Context, decisions, contacts | Unknown | Ongoing |
| (Add your sources) |
Part 2: Repository Health Assessment
| Repository | Currency (1–5) | Completeness (1–5) | Findability (1–5) | Usage (1–5) | Consistency (1–5) | Overall Health |
|---|---|---|---|---|---|---|
| Company Wiki | 2 | 3 | 2 | 3 | 2 | Low |
| Shared Drive (G:) | 3 | 4 | 1 | 4 | 2 | Medium |
| Slack channels | 5 | 1 | 1 | 5 | 3 | Low |
| (Add your sources) |
Part 3: Gap Analysis
| Knowledge Area | Current State | Gap Description | Who Needs It | Impact (H/M/L) | Source of Insight |
|---|---|---|---|---|---|
| New hire onboarding steps | Partial / outdated | No single guide; relies on manager memory | All new hires, HR, managers | H | New hire interviews |
| Client escalation process | Undocumented | Only 2 people know the full process | Customer support, account mgmt | H | Support ticket review |
| Vendor selection criteria | Outdated (2021) | Current criteria not reflected | Procurement, department leads | M | Procurement team interview |
| (Add your gaps) |
Part 4: People Dependencies
| Person / Role | Knowledge Domain | Criticality (H/M/L) | Documented Elsewhere? | Flight Risk | Action Needed |
|---|---|---|---|---|---|
| Senior Ops Manager | Warehouse logistics, vendor contacts | H | Partially (outdated SOP) | Medium (10+ yr tenure) | Capture sessions, update SOP |
| Lead Developer | Legacy system architecture | H | No | High (recruiter interest) | Record walkthrough, pair with junior dev |
| HR Director | Policy exception history | M | No | Low | Decision log creation |
| (Add your dependencies) |
Part 5: Priority Actions
| Priority | Action | Type | Knowledge Area | Owner | Target Date | Status |
|---|---|---|---|---|---|---|
| 1 | Capture legacy system architecture from Lead Developer | Transfer | Engineering | CTO | Not started | |
| 2 | Create single onboarding guide | Create | HR / Onboarding | HR Manager | Not started | |
| 3 | Document client escalation process | Create | Customer Support | Support Lead | Not started | |
| 4 | Consolidate wiki and shared drive policies | Consolidate | Company-wide | Knowledge Lead | Not started | |
| 5 | Archive outdated vendor docs | Retire | Procurement | Procurement Mgr | Not started | |
| (Add your actions) |
Moving Forward
A knowledge audit isn't a one-time event. Your organization's knowledge changes constantly—people join and leave, processes evolve, tools get replaced. The audit gives you a baseline, but you'll need to revisit it periodically to keep your understanding current.
That said, the first audit is the hardest and the most valuable. It takes scattered assumptions and replaces them with evidence. It shows leadership exactly where the risks are. And it gives your team a clear, prioritized plan instead of a vague sense that "we need better documentation."
Start small if you need to. Audit one department. Audit one knowledge area. The process scales, and even a partial audit is infinitely more useful than none at all.
The important thing is to stop guessing and start seeing. Once you can see your knowledge landscape clearly, you can create an internal knowledge base and make it work the way it should—and build an AI knowledge assistant that actually delivers value.
JoySuite can be the destination for your knowledge improvement efforts. Once you've identified what matters and gotten it into shape, Joy makes it accessible—employees ask questions, Joy answers from your content. The work you put into your knowledge audit pays off in findability.
