There's a particular kind of loss that doesn't show up on any balance sheet. Someone who's worked with a client for fifteen years retires, or moves on, and with them goes the unwritten file: this client always wants the executive summary first; that facility's east wing was re-piped in 2009 and the drawings never caught up; the reason you don't schedule the shutdown in March. A new person picks up the account, does everything right, and the client still feels — correctly — that the firm doesn't quite know them anymore.
Everyone agrees this knowledge is valuable. Almost nobody manages to keep it. Here's why, and what actually works.
Why wikis and knowledge bases don't catch it
The standard answer is "document everything in a knowledge base." It rarely works, for a reason worth stating plainly: it depends on people stopping the real work to write things down. Under deadline, that's the first thing to go.
And the most valuable knowledge is the least likely to get written:
- It's tacit. People don't know they know it until the moment it's relevant.
- It's scattered. It lives in email threads, meeting notes, the margins of drawings, and a hundred small decisions — not in one place anyone thinks to document.
- It's "obvious" to the person who has it, right up until they leave and it turns out to have been load-bearing.
A wiki asks the busiest, most experienced people to do extra admin to preserve knowledge they don't perceive as special. That's why the wiki is always out of date.
Capture from the work, not from a form
The shift that makes this tractable: stop asking people to record knowledge, and start building the record from the work that's already happening.
The projects, the correspondence, the documents you produce anyway are full of the details that matter — who asked for what, why a decision went the way it did, the quirks of a particular facility. A quiet record that builds itself from those sources doesn't depend on anyone remembering to fill in a form. There's no wiki to maintain. The institutional memory just accumulates as a byproduct of doing the work.
The other half is retrieval: bringing the relevant details back to whoever's working with that client or facility next — at the moment they need them, not as a 500-page archive nobody reads.
What "good" looks like
You can tell it's working by what stops happening:
- A new person on an old account isn't starting from zero — the history is there when they open the file.
- Nobody has to interrupt the one person who "remembers the Henderson job" to ask what happened on the Henderson job.
- A client doesn't have to re-explain their preferences every time the team changes.
- When someone gives notice, it isn't quietly a small catastrophe.
The measure isn't how much is documented. It's that every project picks up where the last one left off, and the client keeps feeling known even as your people change.
The part you can't compromise on: it stays yours
Institutional knowledge is sensitive precisely because it's valuable — it's the texture of your client relationships. So the rule is firm: your client's information stays yours, and stays in Canada. It isn't shared, and it isn't used to train a model that benefits anyone else. A system that retained your hard-won knowledge by leaking it would be solving the wrong problem.
Where Honewright fits
Client and facility knowledge that stays put is one of the workflows we build for directly: a quiet record that builds itself as you work, picks up the small but important details about each client and facility from the projects you're already doing, and brings the relevant ones back to whoever's working with that client next. No wiki to maintain, no forms to fill.
If your firm's memory walks out the door every time someone retires, tell us what's eating your time and we'll tell you honestly what it would take to keep it.