Why is knowledge sharing so foundational to long-term success with Generative AI (GenAI)?
Because knowledge sharing improves the accuracy and relevance that GenAI needs to generate valuable outputs. The two go hand in hand, especially for enterprises looking to scale the way they gather, create, analyze, and share valuable knowledge.
Our work with Vanguard is a standout example.
Their initiative to elevate the digital employee experience wasn’t just about introducing generative AI — it was about operationalizing it with clear ROI. For Vanguard, the journey began with a commitment to modernizing internal knowledge access while also embedding measurable improvement loops at the core of their digital transformation strategy.
GenAI played a critical dual role.
Before they could deliver real-time, advisory answers to employees and clients alike, Vanguard had to confront fragmented knowledge, unify their internal and external content ecosystems, and implement relevance-tuned AI that could surface highly contextual insights across use cases.
Once that foundation was laid, the result was a transformative ecosystem:
- GenAI experiences personalized for advisors and service agents alike
- Streamlined access to proprietary thought leadership and regulated content
- Elevated customer experiences across both digital and human channels
Vanguard is not alone. Intelligent knowledge delivery is part of the Cisco effort to improve seller productivity. The company uses an AI chatbot that provides sellers descriptive, direct, and relevant answers to questions.
Both stories reveal three truths about knowledge sharing in the age of GenAI.
About Knowledge Sharing and GenAI
Broadly speaking, organizations that prioritize knowledge sharing strategies are better positioned to leverage GenAI effectively. Research shows that many organizations already plan to do so.
According to the 2025 Employee Experience Relevance Report, companies use GenAI in many ways, from customer self-service to knowledge management. Forty-two percent of respondents agreed that their company had invested in generative AI tech and training to help them do their jobs better.
Here’s how knowledge sharing directly supports these GenAI strategies:
1. GenAI Exposes Enterprise Knowledge Gaps and Dirty Data
GenAI models rely on high-quality, comprehensive knowledge bases to generate accurate outputs. When deployed, these systems immediately expose weaknesses in your content, such as:
- Insufficient information: This may prevent GenAI from responding to certain queries at all. This might be due to incomplete documentation, or siloed data sources. For example, the Coveo system uses Relevance Generative Answering; if information is insufficient, the generative answering component will be absent from search results.
- Inconsistent information: When GenAI tries to synthesize contradictory information from different parts of your knowledge base, it may produce confused or conflicting responses.
Knowledge sharing directly addresses these weaknesses, closing the gaps that limit GenAI efficiency.
What’s more, these gaps can be exposed by GenAI systems. According to Dan Shapiro of Organon, a Coveo customer: “GenAI is a funhouse mirror. It will magnify the issues at a level you’ve never experienced before.” GenAI can reveal where critical processes rely on undocumented expertise, or areas where experts haven’t converted their tacit knowledge into explicit, shareable forms.
With the right analytics, you can evaluate when answers are successful for users, or where it’s missing the mark. You can also identify what questions are better off not receiving an answer, like United Airlines’ guardrails around bad actors.
These analytics can also provide confidence scores/user feedback that reveal where GenAI struggles, or usage analytics that show which topics have sufficient knowledge to support GenAI assistance and which don’t.
2. Knowledge Sharing Creates a Cycle of Continuous Improvement
Effective knowledge sharing establishes a self-reinforcing cycle that progressively enhances both your knowledge ecosystem and GenAI capabilities. This cycle operates through several interconnected mechanisms:
- Better Knowledge Exchange → Better AI Outputs
Comprehensive, well-organized knowledge provides GenAI with higher-quality information. Connected knowledge sources enable AI to synthesize insights across domains. And consistent terminology and metadata improve AI’s ability to understand context. - Better AI Outputs → Increased User Trust and Adoption
When users get accurate, relevant responses, they’re more likely to use AI tools. Increased adoption generates more user interactions and feedback. And this engagement provides valuable data on how knowledge is being used. - Increased Trust and Adoption → More Knowledge Creation
As employees see the value of shared knowledge through AI interactions, they become more motivated to contribute. AI can make this process proactive, prompting subject matter experts to address gaps, or automatically capturing knowledge from interactions. - More Knowledge Creation → Enhanced Knowledge Quality
New contributions fill gaps in existing knowledge. User feedback helps refine and correct existing information. It’s a virtuous cycle that supports knowledge management best practices.
Organizations that establish this cycle gain compounding advantages over time, as each improvement to either knowledge sharing or AI capabilities enhances the other.
5 Knowledge Sharing Strategies
Ideally, knowledge sharing tool and knowledge sharing activity is:
- Embedded in employee experience
- Championed by the leadership team
- Enhanced by the power of GenAI
The business impetus is strong. According to the 2025 Commerce Relevance Report, for example, 62% of shoppers are more likely to buy when supported by GenAI-driven guidance. It’s what both customers and employees have come to expect; anything less stands out like a sore thumb.
These five knowledge sharing strategies will help you meet these expectations.
1. Get Real About Knowledge Management Maturity
A maturity assessment will reveal what’s missing from your knowledge sharing practices. You’ll walk away with a priorities list for what to address first. For example:
- Does any of our valuable information live in a knowledge silo?
- Does our knowledge management strategy align with business strategy?
- Do we have metrics in place and the means for accurately tracking them?
- Are we missing critical knowledge sharing opportunities?
In What’s Your Company’s Knowledge Management Maturity Level, I reference the TSIA Knowledge Management Maturity Model 3.0 Framework. APQC also has its own widely referenced maturity model.
The TSIA model evaluates maturity against four core competencies:
- Knowledge sharing culture
- People
- Processes
- Technology
Ultimately, the goal for any enterprise is to get more comprehensive about how it captures, creates, updates, and extends its valuable knowledge assets.
2. Evolve Your Knowledge Management System
The fundamental purpose of any knowledge management system — your internal knowledge base, collaboration tool, etc. — is to support the creation, storage, organization, and sharing of valuable information.
When it comes to the core building blocks of knowledge sharing, knowledge management systems tend to be limited. From content management and search capabilities, to personalization and analytics, an AI search platform makes your knowledge sharing better by:
- Searching, retrieving, and surfacing information from various sources
- Indexing almost any structured or unstructured data type
- Automating content classification, tagging, and organization
- Enabling natural language processing (NLP) and semantic search
- Personalizing experiences based on user history, preferences, and context
- Providing advanced analytics for spotting gaps, patterns, and effectiveness
Together, a knowledge management system and AI search platform not only embed knowledge capture throughout employee workflows, but make sure your teams get the most out of that content.
3. Listen to What the Numbers Say
Analytics and reporting permeate all aspects of mature knowledge sharing practices. The very presence of certain tracking, metrics, and KPIs are indicators of maturity. Not surprisingly, much of this data relates to specific knowledge sharing activities, such as:
- Content creation and participation rates
- Closed incident linking
- Knowledge base gaps identification
- Content archiving frequency
- Knowledge management KPI tracking
Through rich usage analytics, you can learn what helped employees or customers with a similar query and successfully predict what they’ll need next. Machine learning can help you track which resources have proven to be the most helpful, while providing deeper insight into what your people (employees and customers) are looking for.
4. Gain Untapped Subject Matter Expertise
In environments where knowledge sharing is weak, the people with specific expertise are often the same people hoarding it. But knowledge hoarding doesn’t always happen on purpose. On the contrary, many subject matter experts simply lack easy ways to share knowledge.
This is where intelligent swarming can pay dividends.
In intelligent swarming, an employee can connect with expert help for their specific issue. That issue could be as simple as a question. For example, Coveo integrates AI-search with Slack: when a Slack user has a question, they can tap into a unified search index, comprising many different sources of company knowledge, for the answer.
That issue could be a customer service case. Using the same Slack integration, a customer service agent can:
- Initiate a swarm
- Confirm list of subject matter experts
- Swarm with said experts in a dedicated Slack channel
- Supplement the swam with AI-powered search
- Index the closed swarm for use in future searches
Using Coveo, you can embed the same knowledge transfer in your Salesforce workflows.
The point is to make it easy for your subject matter experts — and anyone with relevant knowledge, really — to participate in knowledge transfer. According to KCS Academy, a strong culture of knowledge capture and transfer can improve employee retention by 20-35% (among many other benefits).
5. Scale with Intelligent Knowledge Delivery
You might argue that intelligent swarming is a means for knowledge delivery. You’d be right; but the technology that enables intelligent swarming can do a lot more, at enterprise scale. Indeed, AI-search, machine learning, and passage retrieval allow for two cutting-edge forms of knowledge delivery:
Generative answering answers complex natural-language user queries entered into a Coveo-powered search interface. The answer is generated based solely on the content located in a secure index in your Coveo organization. With generative answering, deliver the right knowledge to:
- Keep users engaged with quick answers
- Guide customers to personalized solutions
- Deflect support cases within the case submission form
- Enable agents with in-the-flow knowledge and insight
- Educate buyers with product learning and discovery
Agentic AI systems act autonomously and make decisions on behalf of the people using them. Agent AI can support scalable, intelligent knowledge discovery through processes that it executes on its own, such as:
- 24/7 customer inquiry resolution
- Prospect engagement
- Sales coaching
- Personal shopping assistance
- Code completion for programmers
Persistent Barriers to Knowledge Sharing
Many organizations still don’t prioritize knowledge sharing best practices, and it’s holding their GenAI deployments back.
The 2025 Knowledge Management Priorities and Trends Survey Report from APQC revealed that 56% of respondents are “just getting started” with or “developing” their knowledge management maturity.
What’s more, four of the APQC’s five knowledge management priorities relate to the way organizations share knowledge:
- Identifying, mapping, or prioritizing critical knowledge (30%)
- Transferring expert knowledge (22%)
- Boosting knowledge management participation or engagement (21%)
- Enabling collaboration across teams/units (20%)
The failure to prioritize knowledge sharing coincides with the persistence of a few other barriers. In How to Build and Promote Organizational Knowledge Sharing, I outline some of the most common barriers to knowledge sharing:
- Knowledge hoarding
- Lack of incentivization
- Difficulties with tacit knowledge capture
- Absence of enterprise-wide knowledge sharing methods
Finally, there’s the issue of scale.
Try as enterprises might to create a knowledge sharing culture, scalability remains a problem. It’s impossible to manually track, manage, and analyze valuable knowledge, let alone deliver the knowledge where and when it’s needed.
With respect to the latter, the Coveo Experience Relevance Maturity Model™ offers some insight into what true scalability requires:
Before you become a leader in relevance maturity — before you tap into the power of GenAI — you need a formal knowledge management process in place. In its 2024 Knowledge Management Maturity Model Survey, TSIA calls effective knowledge management “a precursor to leveraging AI.”
Knowledge Sharing is Just the Beginning
Moving along the maturity scale for knowledge sharing means bringing more intelligence to the entire practice. Thanks to new developments in AI, what begins as an investment in knowledge management can soon expand into new areas — new opportunities to enableboth employees and customers.
As the likes of Vanguard and Cisco show, companies that find ways to scale knowledge sharing opportunities stand to win big.
Dig Deeper
Winning enterprises understand that AI-relevance is a competitive differentiator. Watch our latest Relevance360 installment, Real Customer-First Experiences Start with AI‑Relevance, for expert insight into the five key AI-relevance principles.