Hoarding knowledge: A failure with real consequences
Editor’s note: Kate Woodward is COO of Sharpr, a knowledge management platform based in Chicago. She joined the company as employee nine and has spent over a decade at the intersection of SaaS operations and the insights industry. Find Woodward on LinkedIn.
In May, Pope Leo XIV published his first major encyclical. It runs 245 paragraphs and covers artificial intelligence, human dignity, labor, war and social justice. I am not religious, and I would not normally be reading papal documents. But someone mentioned it had a knowledge management angle, and that was enough to get me in.
What the encyclical actually says
Buried in “Magnifica Humanitas” is an argument that anyone who has spent time in an insights organization will recognize immediately.
The problem identified in the document is structural: Critical knowledge concentrated in too few hands creates imbalance. People make worse decisions, do redundant work and lose the ability to contribute fully – not because they lack intelligence or effort, but because they lack access to what already exists.
His solution draws on the Old Testament book of Nehemiah. After the Babylonian exile, Jerusalem lies in ruins. Nehemiah doesn't rebuild it by centralizing all decisions under one authority. He distributes the work. Every family gets a section of the wall. Everyone has what they need to contribute. The city gets rebuilt because knowledge and responsibility flow to the right people at the right time.
The cautionary tale is Babel. A single language, a single direction, one massive, centralized project. Sound familiar? Impressive on paper, but it collapsed under its own weight.
His point is simple: Hoarding knowledge – whether intentionally or structurally – is a failure with real consequences for the people who needed it and never got it.
When intelligence lives in a folder
I've spent over a decade working with enterprise research and insights teams. And the “Nehemiah problem” is everywhere.
A brand team commissions consumer research, presents it at a quarterly review and files the deck. Six months later, a product team commissions nearly identical research – because they had no way to know the first study existed. The insights were there; they just never moved.
A senior leader makes a strategic call without the market intelligence that would have changed the answer. Not because the intelligence didn't exist. Because it lives in a folder that required knowing exactly what to search for, and the decision couldn't wait.
An analyst spends three days synthesizing category trends for a new business pitch. The synthesis already exists, built by a colleague for a different client. Neither of them knew.
This is the organizational cost of a pull model: knowledge sits and waits. People who need it have to know it exists, remember to look for it, make time to search for it and ask the right question. Most of the time, at least one condition fails.
Why AI doesn't fix the knowledge management problem
The current conversation about AI in knowledge management tends to focus on retrieval. Large language models are genuinely impressive at answering questions – if you ask them. Natural language search, semantic understanding and AI-generated summaries all represent real progress over keyword search.
But they're still part of the pull model.
An LLM sitting on top of your research library is a faster, smarter library. The fundamental dynamic hasn't changed: knowledge waits, and the person has to go get it. They still need to know a question is worth asking. They still need to make time to ask it. The stakeholder who would benefit most from a piece of research (the one who doesn't know what they don't know) still won't find it.
But Nehemiah didn't build a better archive. He sent people to the wall.
The push imperative
The organizations where knowledge actually works aren’t necessarily the ones with the most sophisticated technology. They’re the ones that stopped waiting for people to come looking and started getting the right insights to the right people before the moment passed.
That means thinking about knowledge delivery, not just knowledge storage. It means asking not just, "Can someone find this?" but "Will the person who needs this actually see it?"
In practice it looks like this: A newsletter that surfaces relevant research to stakeholders who would never log into a platform. An alert that reaches a product manager the moment a relevant study lands. An insight brief that travels with an internal email, so every communication carries a live window into the latest findings. Knowledge that moves toward people instead of waiting for people to move toward it.
It also means rethinking what AI is for in this context. The question isn't only, "How do I help someone find things faster when they search?" It's "How do I make sure the right insight reaches the right person even when they're not searching?" Those are different design problems, and most of the industry is only solving for the first one.
Does your team have the knowledge and tools to move forward?
Pope Leo XIV wrote his encyclical about artificial intelligence, human dignity and the future of society. Knowledge management is a small thread in the very large tapestry. But the principle holds at every scale: Knowledge that doesn’t move doesn’t matter.
The goal isn’t a better archive. It’s a better-informed organization. An organization where the right people have what they need, when they need it, without having to go looking for it.
Nehemiah didn't win because he had better stone. He won because everyone knew what they were building and had what they needed to build it.
That's still the job.