If your people aren't clear, how will your agents be?

Brian Chesky didn't save Airbnb with a strategy deck. He asked one question: "What would it take to make every guest feel like they belong?" That question became a compass. It resolved a thousand ambiguous decisions — pricing, host standards, brand voice, market expansion — without needing a thousand policies.

Brian Niccol arrived at Starbucks in 2024 and his first act was to recover exactly that kind of clarity — declaring that the company needed to go "back to Starbucks," back to the coffeehouse, the handcrafted drink, the personal connection — because without it, the stock had lost a third of its value. One leader built a company around a phrase. The other was hired to find the phrase a company had lost.

These aren't slogans. They're compressed identity — a prime directive, a single phrase that tells everyone in the organization what the right answer looks like when the policy manual runs out.

Leaders have always known this matters. The best ones do it instinctively. The question has always been: how much does it actually matter? How do you measure the difference between having this clarity and not having it?

We found a way to test it.

The experiment

We ran 225 decision scenarios across four industries — a consumer brand, a B2B SaaS company, a nonprofit, and a healthcare practice. Each faced five genuinely ambiguous business decisions: the kind with no right answer in the policy manual, the kind that reveal what an organization actually believes.

We tested two configurations against a baseline:

• Baseline — standard organizational context with values and identity
• With a prime directive — the same foundation, plus a single compressed phrase encoding the organization's core identity
• With a bad prime directive — a generic phrase, the kind a strategy offsite produces in an afternoon

Nine independent evaluators scored every output blind, across three dimensions: Decisiveness (does it commit?), Alignment (does it decide as this organization?), and Resonance (would a stakeholder recognize this as "us"?).

What we found

The right phrase changes everything.

A well-crafted prime directive — six words or fewer, encoding the organization's deepest identity — produced measurably better decisions across every industry we tested.

On Alignment specifically — does the organization's decision-maker act as this organization — the prime directive scored 9.00 versus 8.45 at baseline. In healthcare, it reached 9.47, the highest single score in the dataset.

The pattern held across all four industries. The prime directive didn't add information. It added clarity.

A bad prime directive is worse than nothing.

This is the finding that changes the conversation.

We tested a deliberately weak prime directive for one company. Generic. Corporate. The kind of phrase that sounds right in a boardroom but means nothing in a hallway: "Be the best analytics platform."

The bad prime scored lower than baseline. It actively displaced clearer thinking — decision-makers cited the hollow phrase instead of engaging with the real organizational values underneath. A slogan masquerading as a compass doesn't just fail to help. It makes things worse.

Both evaluators flagged this independently. The bad prime "functions more as a slogan than an internalized principle."

Not every organization needs one. That's the diagnostic.

One company — a youth mentorship nonprofit — scored higher at baseline than with the prime directive. Both evaluators agreed independently.

Why? Their values were already so sharp and specific that adding a single lens sometimes narrowed the reasoning. The baseline drew on different principles for different situations. The prime version occasionally went formulaic.

The question isn't "do you need a prime directive?" It's: "When a hard decision arrives, do your values already point clearly in one direction?" If yes, you may not need one. If no — and for most organizations, the answer is no — the prime directive is what breaks the tie.

Knowing which situation you're in is the diagnostic that matters.

Why this matters now

Leaders have always known that organizational clarity matters. The best companies have always had it — compressed into phrases that propagate on their own, shaping decisions long after the founder leaves the room.

What's new is that this clarity is now measurable — and deployable. As organizations begin embedding AI into their operations, the same identity that guides your people needs to guide your agents. An agent constitution without organizational clarity produces decisions that are technically correct and emotionally wrong. The gap between "follows the rules" and "sounds like us" is exactly the gap a prime directive closes.

But this isn't an AI problem. It's a leadership problem that AI makes visible. The CEO who can't articulate the company's identity in six words has the same problem whether they're managing 50 people or 50 agents. AI just makes the cost of that ambiguity measurable for the first time.

What Numen does

Numen is a naming and myth-making practice. The work is compression: finding the six-or-fewer words that encode an organization's identity so precisely that every ambiguous decision resolves against them.

Not a tagline. Not a mission statement. A resolution function — one phrase that collapses ambiguity into clarity.

The data in this paper says phrase quality is everything. The difference between +0.88 lift and +0.11 is the difference between a phrase that resolves decisions and one that replaces them with theater. Getting to the right phrase is craft — synthesis, pressure-testing, narrative architecture. Getting to the wrong one is what most strategy offsites already produce. Consider what that gap costs: an organization makes hundreds of consequential decisions a year — hiring calls, product trade-offs, customer escalations, brand moments. Each one made slightly less aligned, slightly less recognizable as you, compounds. One misaligned brand decision can cost more in recovery than the consulting engagement that would have prevented it. The question isn't whether the phrase is worth crafting. It's whether you can afford to get it wrong.

Numen exists because the difference is measurable and the stakes are no longer abstract.