Last time I wrote about organizational inheritance: practices running on autopilot, reasons lost, the small pancakes on the large pan. The company doing things it cannot explain because the reasoning was never recorded.
There is a level above that, and most mid-sized companies live at it. The reasoning was recorded. Somewhere. Once. By someone. In some format. The company is not running on inheritance. It is running on knowledge.
The problem is that the knowledge requires archaeology to access.
What organizational knowledge looks like
When the person who left in March resigned, not everything she took was tribal. Some of what the company needed was written down. It just was not findable.
The Vizag distributor's ninety-day credit terms? Those were in an email thread from the sales director to the CFO. Fourteen messages deep. Subject line "Re: Re: Vizag update." The approval was real. It was timestamped. It existed. Finding it took the finance team forty-five minutes of searching, a call to the sales director to confirm which thread it was in, and a scroll through a conversation that covered six topics in a single chain.
The vendor volume clause that explained the invoice gap? That was referenced in a comment on a procurement ticket in Jira. The ticket had been closed eighteen months ago. The comment was the fourth in a thread of seven. The comment said "confirmed verbal agreement on volume adjustment, see attached." The attachment was a scanned note that had been uploaded to the wrong ticket and was later moved, and the link in the comment pointed to nothing.
The rationale for the quarterly packaging breakout? That was in a slide, in a deck, from a quarterly review two years ago. The deck was on the former operations head's laptop. When IT reimaged the laptop after he left, the deck was copied to a shared drive, in a folder called "QBR Archives 2022," which contained forty-seven other decks with no naming convention.
Each of these is knowledge. Real, written, timestamped. And each required an excavation to access. The knowledge exists. The knowledge is not useful.
The three costs of archaeology
The first cost is time. Every retrieval takes longer than it should. Twenty minutes is a good outcome. An hour is common. Three hours is not unusual for a cross-departmental decision that touched multiple tools. Multiply that by the number of times someone in a company of two hundred people needs to find a past decision in a given week, and the total is a full-time person's worth of senior hours, every week, spent not on making decisions but on proving that old ones happened.
The second cost is trust. When you finally find the decision, can you rely on it? The email thread shows an approval. Was it the final version, or did a later message revise it? The Jira comment references a verbal agreement. Is the comment accurate? Could anyone have edited it after the fact? The slide in the 2022 deck shows the rationale. Is it the version that was presented, or a draft? None of these tools were designed to prove a decision. They were designed to send a message, track a task, or present a slide. The evidence they contain is incidental, not architectural, and on the day it needs to stand up to scrutiny, it may not hold.
The third cost is the map problem. The knowledge of where the knowledge lives is itself tribal. She knew which thread the Vizag terms were in. She knew which Jira ticket had the vendor clause. She knew which QBR deck had the packaging rationale. When she left, the records did not disappear. But the map to them did. The company now has knowledge it cannot navigate. An archive with no index. A library with no librarian.
Why companies stay at this level
Organizational knowledge feels like a solution. The company has moved past inheritance. Things are written down. There are tools. There are records. When the auditor asks, there is something to produce, even if producing it takes a week.
That feeling is exactly the trap.
Because the knowledge exists, nobody feels the urgency to make it accessible. The problem is not dramatic enough to fix. The email is there if you look hard enough. The ticket is there if you know the right keywords. The deck is there if you remember the quarter. The cost is absorbed into salaries, spread across dozens of people, and experienced as the ordinary friction of work rather than the extraordinary waste it actually is.
Companies build enormous knowledge surfaces without ever building knowledge infrastructure. They have thousands of emails, hundreds of tickets, dozens of drives, and not a single place where a decision can be found by the person who needs it, in the moment they need it, in a form they can trust.
The inbox is not a decision system. It is a communication tool that happens to contain decisions. Jira is not a decision system. It is a task tracker that happens to have approval comments. The ERP is not a decision system. It is a transaction processor that happens to record who clicked "approve." Each tool captures its slice, and the decision, the complete act of someone with authority committing the company on a specific basis with specific reasoning, is scattered across all of them, whole in none.
The distinction that matters
Organizational inheritance cannot answer the question "why do we do this?" at all. The reasoning was never recorded.
Organizational knowledge can answer it, eventually. With effort. If the right mailbox is still active. If somebody remembers which tool it is in. If the comment has not been edited. If the attachment still works. If the deck has not been archived into a folder nobody opens.
"Eventually, with effort, if" is not a decision infrastructure. It is a workaround that degrades with every passing month, every departure, every mailbox purge, every tool migration. The knowledge is real today. In eighteen months, some of it will be gone. Not because anyone deleted it deliberately, but because the tools it lives in have retention policies, the people who knew where it was have moved on, and the fragments that remain are too scattered to reassemble.
A company at this level is not inheriting blindly. It is knowing, precariously. The reasoning exists, somewhere, in some form, accessible to someone who knows where to look and has the time to dig. The next post looks at what happens when a company moves from knowing to being genuinely intelligent, when the decisions are not just recorded somewhere but kept in a way that they can be found in seconds, trusted without verification, and used as input to the next decision without archaeology, without effort, and without depending on whether the person who filed them is still in the building.
Part of The Other Books, an ongoing series on the decisions companies forget.
