By now you want to go find yours. Good. But there’s a right way to read Shadow Systems™, and it starts with the thing your instincts will resist most.
Amnesty. Full, explicit, said-out-loud amnesty.
Because here’s the physics: your people built these systems inside a contract that punishes honesty — we spent a whole chapter on that contract. If the discovery process carries even a whiff of enforcement, you will not find the shadow systems. You will find the ones people were willing to sacrifice, while the load-bearing ones go deeper underground, and your map will be a decoy. The survey lies at exactly the rate the environment makes lying rational.
So you say it plainly, and you mean it: *nothing anyone shows me gets anyone in trouble. I’m not hunting violations. I’m reading evidence.* And then — this is the part that determines everything — the first person who reveals a workaround has to visibly not regret it. One punished confession ends the archaeology forever.
—
Then you ask questions like a field researcher, not an auditor.
Not “are you following the process?” — that question answers itself, falsely. Instead: *Walk me through how this actually gets done. Where do you look when you don’t trust the number? What would break if your laptop died tonight? What do you check before you believe the CRM?* And my personal favorite, which has never once failed to produce a revelation: *what do you do in the gap between how we say this works and how it works?*
Different people, different angles, the same wire-tripping attention to answers that don’t match the question. This is what PBJ Sessions™ are built for — leadership’s version and the team’s version in the same room, where the gaps between them become visible in real time and nobody gets blamed for standing in one.
—
Every shadow system you find, you read the same four ways.
**What job does it do?** Not what rule does it break — what work does it accomplish. There’s always a job. Nobody maintains a workaround recreationally.
**What failure does it mark?** Each one indicts something specific. The reconciliation spreadsheet marks two systems that don’t agree. The private follow-up list marks a CRM nobody trusts. The unofficial Slack channel marks a tool that doesn’t match the work. The workaround points at the wound.
**Who carries it?** Every shadow system has a human paying its maintenance cost — usually one of the absorbers from Chapter 5, usually invisibly. Reading the system means finding the person, and finding the person means finding load you didn’t know you were assigning.
**What breaks if it dies?** This is the triage question. Some shadow systems can retire quietly. Some are genuine liabilities — the ones holding client data in personal accounts need governance conversations, not just gratitude; the full taxonomy lives on the blog. But some are load-bearing, and this rule is absolute: never delete a shadow system before you’ve replaced its function. Every one of them does a job. Kill it before replacing the job, and the job doesn’t disappear — it moves somewhere darker, and this time it knows to hide from you.
Then the promotion path, for the ones that earned it: the shadow system is a working prototype, tested in production by the people who needed it most. Build the official version around what it got right. The people who built the workaround aren’t your compliance problem. They’re your best systems designers, working unpaid, in secret. Hire them for the redesign.
—
Here’s what you actually have, once the map is drawn.
Every shadow system is a piece of the second company — the real one, the one from the very beginning of this manifesto, the one that does all the work while the official company gets all the documentation. You’ve spent this whole book being told it exists. The shadow systems are where it becomes visible: its tools, its pathways, its actual shape, drawn by the people who live in it.
You wanted to meet the second company.
This is its handwriting.