How do we trust a model when the assumptions might be wrong?

You built something that looks precise. But underneath, it runs on guesses — about weather, about how people will use the building, about equipment ten years from now. The unease you feel is not weakness. It is you, seeing clearly. Here is the trap: a polished model invites people to stop questioning it. Including you. You can let the output's confidence stand in for your own, and call that done. That is self-deception, and it will cost you later. What you control is smaller than the model's reach, but it is real. You choose which assumptions go in. You choose whether your team argues them out loud before the client sees a single chart. You choose what you say when someone asks, "how sure are you?" Start there.

◆ How this problem reads on the two dials

GuidanceKnowledge

More coaching

Some to learn

1:1 with AureliusWith others (a Pod)

Some one-to-one

Practise with peers

The team already has the technical skill; what they lack is a disciplined practice for surfacing and owning assumptions together, which is coaching, not teaching.

How the two dials adapt to you →

What’s really going on

You don't trust the model. You own the assumptions. A model is not the future; it is your guesses, wearing numbers. What's in your power is not certainty — it's which assumptions you choose, how plainly you state them, and whether your team tests them before anyone else relies on them.

🔒 What you’ll build togetherUnlock by starting

A moveList every assumption on one page before your next review. If it isn't written down, it isn't being tested.

A moveRun the model twice — best case and worst case — and show both numbers, not just one.

A movePick one person each cycle to argue against your own assumptions, out loud, before anyone else sees the work.

A moveTell the client plainly which numbers are measured and which are guessed. Do not let the formatting blur that line.

A moveAgree now on a trigger point — if real data drifts past a set amount, the team revisits the model together, no debate about whether to.

PractiseAssumption Audit · a Pod of 4 · 30 min

What changes unlock by starting

Your team can name its three riskiest assumptions without checking notes.
You present a range and can defend both ends of it, not just the middle.
Your team catches a bad assumption before a client or regulator does.
The dread of being 'found out' is replaced by a process you trust.

One object, two jobs: a public answer to a real problem, and — the moment you start the chat — Aurelius’s live plan for your version of it.