Where your company's knowledge actually lives

Your operations manager resigns on a Friday. On Monday you find out the approval rules for your three largest accounts lived in her head and in one chat thread nobody else was in.

That is not a staffing problem. It is a knowledge problem, and nearly every business between 2 and 25 people has it.

Most owners have never asked where their company knowledge is stored, because the question sounds like it has an obvious answer. It does not. The honest answer is usually everywhere and nowhere.

The knowledge in your business has no address¶

Ask four people on your team where the current refund policy lives. You will get four answers, and at least one will be a name rather than a location.

That is the test. If the answer to "where is it" is a person, the knowledge is not stored. It is borrowed.

Nothing about this feels urgent, which is exactly why it stays broken for years.

What scattered information costs a 12-person team¶

A 12-person insurance brokerage had four people quoting commercial policies. Each kept private carrier notes in their own files. Two of them left within five weeks of each other. Quote turnaround went from one day to four, and it took eleven weeks to rebuild what those notes held.

Nobody logged that as a loss. It showed up as slow quarters and frustrated clients.

The cost of this pattern is always paid twice. Once when someone leaves, and again every time a new hire spends six weeks learning what should have taken two days to read.

The four places knowledge hides in a small business¶

• Chat threads. Decisions get made in a conversation, then scroll out of reach.
• Inboxes. The final version of the contract terms sits in one person's sent folder.
• Shared drives. Files exist, but three versions do, and nobody knows which is current.
• People's heads. The largest store by volume, and the only one that can quit.

Every business has all four. The difference is whether you know it.

How to map your knowledge in one afternoon¶

1. List your ten most repeated processes. Quoting, onboarding, invoicing, whatever runs weekly.
2. For each one, write down where the instructions live right now.
3. Mark every process whose answer is a person's name.
4. Count them. That number is your exposure.
5. Pick the three with the highest cost if that person is unavailable for a month.

You now have a queue. Start writing those three down before anything else.

Why organized information is the prerequisite for useful AI¶

An AI system reads what you give it. If your pricing rules exist only in a conversation from March, the model cannot use them, and neither can your new hire.

This is the part owners skip. They buy a tool, point it at a messy drive, and get vague answers back. The tool is not the problem. The input is.

Organized information is cheap to build and expensive to retrofit.

What to do from here¶

Do not start with software. Start with the count from step four above.

Write down the three processes that would hurt most to lose. Put them somewhere with an address that is not a person. Then look at automation.

Knowledge you cannot point to is knowledge you do not own.

Keep exploring¶

If you want a clear baseline of where your process knowledge sits before you automate anything, start the AI Readiness Audit or contact FIT.