> ## Content Index > Fetch the complete content index at: https://jonathancanales.com/llms.txt > Use this file to discover other available public pages before exploring further. # Your company's knowledge isn't worth much if it lives locked inside departments - URL: https://jonathancanales.com/en/knowledge-locked-away/ - Published: 2026-10-22T14:00:00.000Z - Updated: 2026-10-22T21:23:02.000Z - Description: Quick question: if a client asks you something that touches sales, operations, and finance, how long does your company take to give a complete answer? - Author: Jonathan Canales - Tags: Leadership and teams, #en Your company's knowledge isn't worth much if it lives locked inside departments 0:00 /434 1× [![Listen on Spotify](https://jonathancanales.com/content/images/2026/10/image-ebade917.jpg)](https://open.spotify.com/show/1RbZ5C2DI8VFR4RA6paicv?ref=jonathancanales.com) [Read in Spanish →](https://jonathancanales.com/conocimiento-encerrado/) Quick question: if a client asks you something that touches sales, operations, and finance, how long does your company take to give a complete answer? If the answer is "depends on three people replying to the email," you have the problem described by the fourth cultural shift in the Harvard Business Review article this series is based on: optimizing each area on its own instead of thinking like one company. ## The prize for the area, the penalty for the company The article describes the default of almost every organization: people are rewarded for optimizing what they control. Their function, their budget, their process. Sales sells, operations delivers, finance guards the money, and everyone defends their square meter. It works reasonably well until something arrives that crosses the borders, and AI is exactly that: its biggest value shows up when it connects knowledge that lives in different areas. ![Person celebrating with a trophy with an orange star](https://jonathancanales.com/content/images/2026/10/hbr5-d1-g.png) And here the article adds a serious warning: when people feel their job is threatened, they do the opposite of connecting. They protect information, defend their expertise, and retreat into local priorities. AI arrives offering to share everyone's knowledge and, right at that moment, fear turns each area into a fiefdom with the door shut. If you don't work the [culture](https://jonathancanales.com/en/ai-ready-culture/), the technology lands on frozen ground. ## Six repositories, forty hours The most illustrative case in the article comes from a semiconductor manufacturer. Resolving an issue required experts to search separately across six different functional repositories, routinely consuming more than 40 hours. When they brought in AI agents, that work dropped to less than an hour, with better completeness. The key phrase: they stopped having six functional versions of the problem and gained a single enterprise view. ![Stack of folders with one orange folder](https://jonathancanales.com/content/images/2026/10/hbr5-d2-g.png) Think about how much of that happens in your business with no semiconductors involved. The customer's history lives in the sales inbox, complaints in WhatsApp, payments in the accounting system, and project details in somebody's head. Every complete answer requires a pilgrimage. AI can join all that, but only if the information exists outside of heads and fiefdoms. ## The hidden cost of local optimization There's a detail almost nobody adds up: local optimization has costs that show up in nobody's budget. When sales optimizes for closing, operations inherits impossible commitments; when purchasing optimizes for price, production inherits defective material; when every area wins its metric, the company loses the result. The article frames it as an adaptation risk: the most adaptable organizations are the ones that connect work across internal borders, because new problems (and AI is one of them) never respect the org chart. ## Technology connects data; leaders connect people The second case in the article feels even more important to me because it's purely human. A company struggling with silos across hundreds of group companies placed leaders on cross-regional, cross-functional teams and challenged them to solve shared business problems using resources and expertise from across the enterprise. The experience exposed people to capabilities and colleagues they didn't know existed and helped build a "one company" mindset. Within a year, perceptions of trust, transparency, inclusion, and collaboration rose by an average of three percentage points globally. The authors sum it up in a line that should be a poster: AI can connect the knowledge; leaders have to connect the organization. The tool joins data. Trust between areas gets built by people, through shared projects, common goals, and recognition for whoever helps beyond their square meter. ## Fiefdoms, small-business edition In a small company the silos aren't departments: they're people. "Ask Lupe about that" is a silo. "Only Memo has that file" is a silo. "Sales doesn't tell operations until after signing" is a silo with consequences. The small-business version of enterprise thinking looks like this: - Important information lives in searchable places, not in heads or personal chats. - Problems that touch several areas have a shared owner, not a chain of "that's not mine." - The one who helps another area gets recognized, not just the one who hits their number. And an honest warning: if your people sense that AI is coming to make someone surplus, nobody is going to document what they know. Documenting your expertise becomes handing over the keys to your job. That's why the culture piece comes first: people share when sharing benefits them, not when it threatens them. ## A five-minute test If you want to know how connected your company really is, run this thought experiment: an important client calls with an urgent issue that touches three areas, and you're on vacation with no signal. What happens? If the answer is "they sort it out together because they know each other, trust each other, and have the information," congratulations: you have enterprise thinking. If the answer is "they wait for you to come back" or "everyone answers their piece and the client assembles the puzzle," you know where the work is. AI can give you shared information; the trust to use it without you gets built beforehand. ![Stopwatch with an orange hand](https://jonathancanales.com/content/images/2026/10/hbr5-d3-g.png) ## What to do this week - Pick ONE topic that today requires asking two or more people (the status of a client, for example) and pull its information into a single searchable place. That's your first "enterprise repository." - Identify what Lupe and Memo know that only they know, and start getting it into documents. No drama: half an hour a week. - Check your incentives: is there any reward, even informal, for helping another area? If not, don't expect spontaneous collaboration. ## Questions to ask yourself Which customer questions take days because they touch several areas? What critical information disappears if one person is out tomorrow? Do my people believe sharing what they know makes them more valuable or more replaceable? ## The takeaway AI makes the most of connected knowledge and starves on locked-up knowledge. You can buy the best tool in the world: if your information lives in fiefdoms and your people believe sharing puts them at risk, you bought plumbing for a house with no water. Connect the company first; then AI does magic with it. Source: "There's No Such Thing as an AI-Ready Culture", Sarah Jensen Clayton, Michael Welch and Khoi Tu (Korn Ferry), Harvard Business Review, October 5, 2026\. [https://hbr.org/2026/10/theres-no-such-thing-as-an-ai-ready-culture](https://hbr.org/2026/10/theres-no-such-thing-as-an-ai-ready-culture?ref=jonathancanales.com)