The method

How a business remembers.

Arc is not a place to put documents. It is one loop, running continuously: hold what the company believes, watch what changes, explain what it means, act only with your approval, and compare what happened against what you expected.

The loop

Five moves, running continuously.

Most tools do one of these. A wiki remembers without noticing. A dashboard notices without remembering. An AI chat judges without either. The value is in closing the loop.

Remember

Goals, context, evidence, decisions, rationale, assumptions and outcomes — kept together, not scattered across tools that each hold one fragment.

Notice

Monitors watch the company, customer, market and competitor changes that would matter, and name which part of the strategy each one touches.

Judge

What changed, which beliefs it affects, how confident Arc is, and which options deserve consideration — with the reasoning visible.

Act

Approved decisions become bounded work. Nothing consequential happens without you saying so.

Learn

What you expected, set against what actually happened, so the company’s judgment improves instead of resetting.

What Arc holds

Eight objects, and nothing loose.

Arc’s vocabulary is deliberately small. Everything the company knows lands in one of these, which is what makes it possible to ask a question later and get a real answer rather than a search result.

Thesis
What the business currently believes about how it will create value. Living, versioned, and answerable to evidence.
Decision
A choice with its rationale, evidence, owner, expected result and review date attached.
Assumption
A condition that must stay true for the strategy to work — tracked, and flagged when evidence starts contradicting it.
Evidence
A sourced observation that supports or weakens a belief, labelled for what it actually is: fact, estimate, belief or inference.
Signal
A meaningful change, inside the company or outside it.
Monitor
A bounded process that watches one signal, assumption or responsibility.
Brief
A short weekly synthesis: what changed, what it affects, what deserves your attention.
Timeline
The chronological view of decisions, changes and outcomes — where the strategy actually moved, and why.

Getting context in

The reason most decision logs die.

Every owner who has tried this has the same story. You start a decisions doc. You keep it for six weeks. Then a hard month arrives and writing it up is the first thing to fall off the list. Six months later the doc is a museum of a business that no longer exists.

Arc treats that as the central design problem rather than a discipline problem. If keeping the memory current depends on you remembering to keep it current, it will not survive contact with a real week.

You explain the business once. After that, Arc learns from the work you were already doing.

The first pass is a guided interview — Arc asks the high-value questions and leaves the rest alone. From there it takes what you already produce: the documents you were going to write anyway, the calls you were going to have, the numbers your tools already hold. When something looks like a decision or an assumption, Arc drafts the entry and asks you to confirm it. You approve, correct, or ignore.

Nothing is saved to your Arc because a model was confident. It is saved because you said yes.

Judgment you can inspect

Six questions Arc will always answer.

A recommendation you cannot interrogate is just a confident sentence. These work on anything in your Arc, at any time.

  • Why do you believe this?
  • What changed?
  • What evidence contradicts it?
  • What would change your mind?
  • When should we revisit this?
  • What happened the last time we made a similar decision?

See whether your business is a fit.

Founding Arc is a small design-partner group of owner-operated companies. If the strategy still lives in your head, that is the profile.

Request early access

Founding Arc is a private design-partner program. A conversation, not a checkout.