Last week, eGain unveiled the findings of a recent study in partnership with IDC, The Knowledge Foundation That Unlocks Enterprise AI Success, based on a survey of 520 organizations. Findings show that the return you get from AI has less to do with the model you pick, with more impact coming from the knowledge you give it to work with.

Organizations with the most mature knowledge management practices report roughly twice the business impact from AI as those with the least mature practices.

If your leadership team has asked what AI has delivered so far, you’ve probably seen the answers vary from team to team. Some teams can point to metrics such as faster resolution time, while others do not have clear, measured results yet.

This article walks through the key findings from this survey, exploring the pillars that explain why and how investments in AI pay off.

The Three Pillars of a Knowledge Foundation: Technology, Process, People and Structure

IDC grouped organizations by how mature their knowledge management practices are and compared the most mature, “Leaders”, with the least mature, which it calls “Laggards”. Leaders reported a significant improvement on every business impact metric in the study, from ticket deflection to customer satisfaction.

According to IDC, the difference comes down to the knowledge foundation. The foundation is incorporated with a mix of technology, process, and people that keeps an organization’s knowledge accurate, connected, and ready for both employees and AI to use.

Technology gives people and AI one source of truth

The first pillar is a unified knowledge platform, which gives employees and AI agents a single source of truth to draw from. When everyone works from the same governed content, a customer receives consistent answers no matter which channel they refer to, such as a speaking to a live agent or using a chatbot.

Laggards tend to keep relying on document management systems and collaboration tools, which work well for storing files but weren’t designed to give AI reliable answers. You can read more about repositories vs. Knowledge management systems here.

Process keeps knowledge current as it grows

The process pillar covers the ongoing work of keeping knowledge current and making sure it moves between departments, so articles don’t go stale a few months after someone publishes them.

At enterprise scale, automation is commonly used to maintain a knowledge base. This can include capturing the latest content, tagging it, routing it to the right people, and generating answers, so the knowledge base stays up to date without someone reviewing every article by hand.

People and structure decide who owns knowledge

The third pillar is about ownership and skills. Knowledge tends to fall out of structure when nobody is appointed responsible for it, and many organizations leave each department to manage its own content in its own way.

Leaders take a more centralized approach, with a clear mandate for those who own knowledge across the business. IDC also found that the personal leading knowledge management looks noticeably different at leader organizations, which makes it easier to connect knowledge decisions to the company’s broader AI plans.

Knowledge engineers are increasingly valuable as they build the shared vocabulary and structure that help AI understand how information relates across the business. Leaders are much more likely to have people with those skills on their teams.

How COUNTRY Financial improved results with a stronger knowledge foundation

COUNTRY Financial is an insurance and financial services company that serves 1.5 million policyholders across 19 states.

Its roughly 1,700 agents must know the details of policies across states with different regulations, often while a client is sitting across the desk. COUNTRY set out to streamline agents’ workflows, save time, and reduce reliance on underwriters.

COUNTRY launched AI Assist, powered by the eGain AI Knowledge Hub. The assistant gives generative AI answers drawn only from COUNTRY’s own curated content, such as guidelines, underwriting notes, and policies.

After rolling out AI Assist, COUNTRY Financial achieved clear gains:

• 25% fewer underwriting calls in the first month
• 96% answer accuracy
• 72% resolution rate: Reps got the answer they needed without calling anyone
• 70% rep usage rate, with 2,500 unique field users. This was more than the agent count, because associate reps adopted it too.
• 31,000 interactions in the first two months

Where to start building your knowledge foundation

When an AI agent pulls from outdated or conflicting information, it gives wrong answers with confidence and can spread misinformation across multiple channels. Over time, this wears down trust with customers and employees, and in regulated industries it can also create compliance risk.

The organizations seeing the best results avoid this by working on these three pillars. A strong platform without clear ownership, or a skilled team without automation, still leaves gaps that AI will eventually expose.

To dig into the full research, download the IDC white paper, The Knowledge Foundation That Unlocks Enterprise AI Success, which includes more insights on what differentiates leaders from laggers, key capabilities to focus on, and where to invest first.

If you’d like to see where your organization stands today, take eGain’s free AI Content Readiness Assessment.