First Principles: What's Actually Possible?

Loading Company Context

Before starting, look for a company-context.md file. Read it if available, focusing on:

• Current strategy — especially any first-principles reference point, stated theoretical maximum, or gap analysis from a previous session
• Active assumptions — conventional constraints that were identified previously. Are they still assumed? Have any been tested or broken?
• Financial situation — what resources are available for closing the gap? This affects Phase 6 (working backwards to today)
• Session log — when was first-principles last run? What was found?

If context exists with a previous theoretical maximum analysis, you're in Review mode unless the user explicitly asks for a full diagnostic.

If no first-principles analysis exists, you're in Diagnostic mode.

If no context file exists at all, ask:

• What business or problem are we applying first principles to?
• What dimension of performance matters most — cost, speed, quality, scale, something else?
• How long has the current approach been in place, and where did it come from?

Determine the Mode

Ask: "Are we doing a full first-principles diagnostic, reviewing an existing analysis, or setting up alert triggers?"

Then follow the appropriate track below.

Mode 1: Diagnostic (Full Session)

Time guidance: 45+ minutes. This is the deep one. You're dismantling assumptions that may have been load-bearing walls for years. Don't rush it. If a constraint feels obviously real, you probably haven't questioned it hard enough yet.

Before You Begin: Physical vs. Conventional Constraints

Most of what people treat as constraints aren't. They're conventions. Habits. "The way things are done." The difference matters more than almost anything else in strategy.

Physical constraints are things the universe won't let you violate. The speed of light. Thermodynamics. The tensile strength of steel. Biological limits on how fast an organism grows. These are real. You can't argue with physics.

Conventional constraints are things humans decided at some point, for reasons that may or may not still apply. Industry standards. Pricing norms. "Best practices." Organisational structures. Regulatory requirements (which are human decisions, not physics — they can change). The way supply chains have always been configured. The assumption that customers want what they've always wanted.

Here's the problem: most people can't tell the difference. They've been inside the conventions so long that the walls look like bedrock. A fish doesn't know it's in water. A strategist who's spent 20 years in an industry doesn't know which of their assumptions are physics and which are just water.

First-principles thinking means draining the water and seeing what's actually solid underneath.

Elon Musk used this at SpaceX. The conventional constraint said rocket launches cost $60M+. Everyone accepted that because everyone had always accepted that. The physical constraint — the actual cost of the raw materials in a rocket — was about 2% of that number. The other 98% was convention, process, margin stacking, and "the way aerospace works." He didn't accept the conventional constraint. He asked what was physically required, and built from there.

You don't have to be building rockets. The same logic applies to a consulting firm, a manufacturing company, a software business, a local service provider. The question is always the same: what does physics actually require, and what are we just assuming?

Phase 1: Define the Theoretical Maximum

Pick the dimension that matters most. Cost per unit. Speed of delivery. Quality level. Scale of production. Revenue per employee. Customer acquisition cost. Whatever the thing is that, if you could push it to the absolute limit, would transform the business.

Now push it — not to "best in class" or "top quartile." Push it to the theoretical maximum. The physical limit. What would be possible if every conventional constraint were removed and only physics remained?

This feels uncomfortable. Good. If it doesn't feel slightly absurd, you're not pushing hard enough.

The rocket example: The theoretical minimum cost of a rocket is the raw material cost. That's the physical constraint. Everything above that — labour, margin, process overhead, supplier markup — is conventional. It can be questioned, redesigned, or eliminated.

The professional services example: The theoretical maximum revenue per person is bounded by hours in a day (physical) and the value per hour the market will pay (a mix of physical and conventional). But the conventional constraints — the billable hour model, the partner leverage ratio, the assumption that delivery requires senior people's time — those are all choices. A firm that productises its expertise, uses technology to deliver, and decouples revenue from headcount isn't violating physics. It's violating convention.

Ask:

• "What dimension are we optimising? Pick one. If you try to optimise everything, you optimise nothing."
• "What's the physical limit on that dimension? Not the industry benchmark. Not the best competitor. The actual limit imposed by physics, biology, or mathematics."
• "What would it look like to operate at that limit? Describe it — even if it sounds impossible with today's approach."

Don't let the user anchor on current performance. The whole point is to escape the gravitational pull of "how things are." The theoretical maximum is a compass bearing, not a destination. You may never reach it. But if you're not even pointing at it, you're navigating with a broken compass — and every step takes you further from where you could be.

Phase 2: Map the Gap Honestly

Now you have two numbers: where you are and the theoretical maximum. The gap between them is where all the interesting strategic questions live.

Express the gap as a ratio or multiple, not a percentage. Percentages domesticate large gaps. "We're at 15% of theoretical maximum" sounds clinical. "We're 7x away from what's physically possible" hits different. It should hit different. A 7x gap means there's a 7x opportunity sitting inside your existing business, hidden behind assumptions.

Ask:

• "What's our current performance on this dimension?"
• "What's the ratio between current and theoretical maximum?"
• "Is that gap surprising? Smaller or larger than you expected?"

If the gap is small (less than 2x), either the dimension is already well-optimised or the theoretical maximum wasn't pushed hard enough. Go back to Phase 1 and push harder.

If the gap is large (5x+), you're looking at a business that's mostly made of conventions. That's not a criticism — it's an opportunity. Most of that gap is addressable without violating any laws of physics.

Phase 3: Classify What's In the Gap

The gap isn't one thing. It's a stack of different constraints, and they need different strategies. Walk through each layer:

Physical / Economic Constraints

Things that are genuinely hard or impossible to change. Material properties. Energy requirements. Geographic distance. Fundamental unit economics that can't be designed away. These define the floor of what's achievable. Respect them.

Conventional Constraints

"The way it's done." Industry norms, standard operating procedures, assumptions about what customers want, inherited organisational designs. These are the buggy whips — they exist because they've always existed, and entire careers are built around maintaining them. They feel permanent but they're not.

Ask for each one: "If we were starting this business from scratch today, with no legacy, would we choose to do it this way?" If the answer is no, it's conventional, not physical.

Regulatory Constraints

Human-made rules enforced by governments. They're real in the sense that violating them has consequences. But they're not physics. They change. They vary by jurisdiction. They can be influenced, worked around, or anticipated. The question isn't "is there a regulation?" — it's "is this regulation likely to persist, and is there a way to deliver the same outcome within different regulatory frameworks?"

Technological Constraints

Things that aren't possible with current technology but aren't physically impossible. These are temporary walls. The question is timing: is the enabling technology 2 years away, 10 years away, or 50? If it's 2-5 years, you should be planning for it now. If it's 50, note it and move on.

For each constraint in the gap, classify it. The conventional ones are where most of the strategic opportunity sits. The physical ones are where you stop pushing and start accepting. Confusing the two is the most expensive mistake in strategy.

Phase 4: Invert — What Would Guarantee Failure?

Flip the question. Instead of asking "how do we close the gap?" ask: "What would guarantee we never close it? What would ensure we stay exactly where we are, or get worse?"

This is Charlie Munger's inversion principle, and it's devastating when applied honestly. Because the things that guarantee failure are usually the things organisations are already doing.

• Optimising the current process instead of questioning whether it's the right process
• Benchmarking against competitors who are all stuck in the same conventions
• Investing in efficiency improvements to a fundamentally limited model
• Promoting people who are experts in the current way and threatened by alternatives
• Defining success as incremental improvement rather than step-change
• Treating regulatory and conventional constraints as equivalent to physical ones
• Waiting for "proof" that the theoretical maximum is achievable before acting

Ask:

• "Which of these failure-guarantees are we currently doing?"
• "What would an outsider — someone who knows nothing about 'how our industry works' — find absurd about how we operate?"
• "What are we defending that we should be dismantling?"

The things that make you defensive when you read them? Those are the ones to pay attention to.

Phase 5: Set Strategic Intent

Now you know the theoretical maximum, the gap, what's in the gap, and what guarantees failure. Time to set a strategic intent — not a goal, not a target, not a forecast. An intent.

Strategic intent is a 10-25 year orientation. It's the peak you're climbing. It doesn't need to be "realistic" in the sense that you can draw a straight line from here to there. It needs to be physically possible (not violating any genuine constraints) and directionally correct (pointing toward the theoretical maximum, not away from it).

Jim Collins asked the right filter question: "If you could pick only one summit, and you'd spend 10-25 years climbing it, which summit would you pick?" That's strategic intent. It's not a plan. It's a declaration of which mountain you're on.

The danger of not setting strategic intent is that you end up on the wrong mountain. You optimise beautifully for a local peak — you get really, really good at making buggy whips — and then the landscape shifts and your peak becomes irrelevant. Strategic intent keeps you oriented toward the highest peak, even when the immediate path is unclear.

Ask:

• "Given the theoretical maximum and the gap analysis, what's the summit? Not a 3-year target — a 10-25 year direction."
• "Is this physically possible? Does it violate any genuine constraints?"
• "Is this the highest peak, or are we choosing a more comfortable one because the path looks easier?"
• "What would we have to believe for this to be wrong? What assumption, if violated, means we're on the wrong mountain?"

Phase 6: Work Backwards to Today

Strategic intent without a first step is a motivational poster. Now work backwards from the summit to today.

The adjacent possible test: You can't jump from base camp to the summit. You can only move to the next camp — the thing that's achievable from where you currently stand, with the resources you currently have, in a timeframe that keeps momentum alive.

Ask:

• "Standing where we are today, what's the single most important constraint preventing us from moving toward the summit?"
• "Is that constraint physical, conventional, regulatory, or technological?"
• "If it's conventional — what would it take to break it? Not work around it. Break it."
• "If it's physical — what would it take to shift the physical constraint itself? (New materials, new processes, new science?)"

The first intervention: Name one thing. The single most leveraged action you can take in the next 90 days that moves you toward the summit. Not a committee. Not a study. An intervention — something that changes the system.

This is the binding constraint — the one bottleneck that, if released, enables everything behind it to flow. Find it, name it, act on it.

Ask:

• "What's the first intervention? Specifically — what will you do, by when, and how will you know it worked?"
• "Is this intervention attacking a conventional constraint or a physical one? If physical, are you sure?"
• "What happens if it works? What's the second intervention that becomes possible?"

Closing the Diagnostic

No report. No slide deck. Instead, reflect back:

1. Theoretical maximum: Restate it. The physical limit on the dimension that matters most.
2. Gap ratio: Where you are vs. where physics says you could be. Express it as a multiple.
3. Conventional vs. physical: How much of the gap is convention (addressable) vs. physics (hard limit)?
4. Strategic intent: The 10-25 year summit. The mountain you've chosen.
5. First intervention: The one thing, in the next 90 days, that starts the climb.

Then ask: "Does this feel right — or does it feel too ambitious? If it feels comfortable, we haven't pushed hard enough."

A Note on Incremental Goals

Incremental goals are dangerous when they're oriented toward a local peak. Getting 10% better at the current approach feels productive. It feels responsible. It feels like progress. And it is progress — toward the top of the hill you're already on.

The problem is that the hill you're on might not be the mountain. A 10% improvement to buggy whip manufacturing in 1905 was technically progress. It was also strategically irrelevant because the landscape was about to shift permanently.

First-principles thinking doesn't say "don't improve." It says: make sure you're improving in the direction of the theoretical maximum, not just the local maximum. If your incremental improvements are making you better at something the world is about to stop needing, you're polishing the brass on the Titanic.

The compass analogy: incremental improvement without first-principles orientation is walking faster with a broken compass. You cover more ground. You feel productive. And you're getting further from where you need to be with every step.

Connection to Two-Track Model

First-principles analysis naturally produces two tracks of work:

Track One — Exploit the current position: Optimise within the conventional constraints you've chosen to keep (for now). Extract maximum value from the current model. This is where the next 12-24 months of cash flow lives.

Track Two — Build toward the theoretical maximum: Invest in breaking the conventional constraints that stand between you and the summit. This is where the next 5-15 years of competitive advantage lives. It won't show ROI on a quarterly basis. It will look like waste to anyone who doesn't understand the first-principles analysis behind it.

The two tracks are not either/or. They're both/and. Track One funds Track Two. Track Two ensures Track One doesn't become a dead end.

The worst strategic mistake is killing Track Two to improve Track One's quarterly numbers. You're harvesting the seed corn. The second worst is abandoning Track One to chase Track Two before Track Two can sustain itself. You're starving before the harvest.

Update company-context.md

After the diagnostic, update company-context.md with:

Current strategy section — add the first-principles reference point:

• Theoretical maximum (dimension + value or description)
• Current performance and gap ratio
• Summit / strategic intent statement

Active assumptions table — add every conventional constraint identified:

Session log — add a row:

Mode 2: Review

Time guidance: 15-20 minutes. This is a check, not a rebuild. If you start finding that the foundations have shifted, switch to Diagnostic mode — the old analysis can't be trusted anymore.

Steps

1. Load company-context.md. Read the existing first-principles reference point, active assumptions, and session log.

1. Has the theoretical maximum shifted?

• New technology that changes what's physically possible?
• Regulatory change that removes (or adds) a constraint?
• Scientific or engineering breakthrough that moves the ceiling?
• If yes, restate the theoretical maximum and recalculate the gap ratio.

1. Has the gap ratio changed?

• Have you moved closer to the theoretical maximum since last session?
• Has the gap widened (competitors moved, or you regressed)?
• Express the updated ratio. Compare to the previous one.

1. Have any conventional constraints been broken or new ones appeared?

• Walk through the Active assumptions table. For each conventional constraint:
• Is it still assumed? Has anyone tested it?
• Has a competitor broken it? Has a startup ignored it?
• Has a new conventional constraint appeared that nobody's questioning yet?
• Update the table — change confidence levels, add new constraints, mark broken ones.

1. Is the strategic intent still the right peak?

• Given any shifts in theoretical maximum or constraints, is the summit still correct?
• Or has the landscape changed enough that you're now climbing a local peak?
• This is the hardest question in the review. Nobody wants to hear that their mountain moved.

1. Is the first intervention still the binding constraint?

• Did the previous first intervention get executed? What happened?
• If it worked, what's the new binding constraint?
• If it didn't work, why not? Was it the wrong intervention, or was execution the problem?

1. Update timestamps. Every assumption checked gets today's date. Every unchanged item gets confirmed. The trail matters — it makes the next review faster and more honest.

Update company-context.md with all changes. Add a session log entry.

Mode 3: Alert Triggers

These are the tripwires that tell you your first-principles analysis needs revisiting. Set them, then monitor them. When one fires, don't ignore it — that's how companies end up optimising buggy whips while the Model T rolls off the line.

Constraint-Shift Triggers

The constraints you identified in your analysis are the foundation. If any of them move, the whole structure needs checking.

• Regulation changes: A regulation you classified as a constraint gets revised, repealed, or reinterpreted. Or a new regulation appears that creates a constraint you didn't have.
• Technology curves hit inflection points: A technology you classified as "not yet ready" reaches product stage. Or a technology you depend on gets commoditised (which may be an opportunity, not just a threat). Watch for the S-curve — at 7% adoption, most people think it's niche. By the time it's at 25%, it's unstoppable.
• A competitor breaks a conventional constraint: Someone in your market (or an adjacent one) does the thing your industry agreed was impossible. This is the most important trigger. When someone demonstrates that a convention was just a convention, the clock starts ticking for everyone else.

Peak-Shift Triggers

The theoretical maximum itself moved. This is rarer but more consequential.

• A scientific or engineering breakthrough changes what's physically possible
• A new material, process, or platform creates a step-change in the theoretical limit
• A market shift changes which dimension matters most (your theoretical maximum was on cost, but now the market rewards speed — the peak moved)

When the peak shifts, the strategic intent may need to shift with it. This is a full Diagnostic trigger.

Local-Peak Triggers

Signs you're optimising a local peak while the landscape shifts underneath you.

• Your incremental improvements are getting smaller (diminishing returns = approaching the top of a local peak)
• New entrants are approaching the problem from a fundamentally different direction and gaining traction
• Customers start asking for things your current model can't deliver, no matter how well you optimise it
• Your best people start leaving for companies that are organised around a different theoretical maximum
• You're investing more and more to maintain the same position — running faster to stay in place

These are the hardest triggers to act on because the local peak still feels like success. You're still growing. You're still profitable. You're still "winning." But the rate of improvement is declining, and somewhere else, someone is climbing a higher mountain.

Recording Triggers

Add triggers to the Active assumptions table in company-context.md:

When a trigger fires, run either a Review (if it's a single constraint question) or a full Diagnostic (if the landscape has shifted broadly).

Connection to Other Skills

Environmental Radar (eterdis-environmental-radar): External forces are what shift constraints. A PESTEL force that changes regulation, technology availability, or market demand can move a constraint from "physical" to "conventional" or vice versa. Run the Environmental Radar to identify forces, then check each one against your first-principles constraint classification. If a force is acting on a constraint you assumed was stable, that's a trigger.

Wardley Map (eterdis-wardley-map): Component evolution changes what's conventional versus physical. A component you're building custom (treating as a physical constraint on cost or speed) may be evolving toward commodity — meaning the constraint is actually conventional, and someone else will break it by buying what you're building. The Wardley Map shows which components are moving; first principles tells you which movements change the theoretical maximum.

VRIO (eterdis-vrio): The resources needed to close the gap between current performance and the theoretical maximum — those resources need to pass the VRIO test. If closing the gap requires a resource you don't have and can't build (not valuable, not rare, or not organised to exploit), the first intervention needs to address the resource gap before it can address the performance gap.

Market Position / Playing to Win (eterdis-market-position): Strategic intent from first principles feeds directly into the winning aspiration in Playing to Win. The theoretical maximum defines the ceiling of what's possible; Playing to Win defines where to play and how to win within that possibility space. If your winning aspiration is lower than your theoretical maximum, ask why — is it a deliberate choice or an unexamined assumption?

Strategy Map: The first intervention identified in Phase 6 should appear as the primary gap in the strategy map. If it doesn't, either the strategy map is missing the binding constraint or the first-principles analysis identified the wrong one. They should agree. If they don't, figure out why before acting on either.

First-principles strategy as applied through Eterdis consulting practice. This isn't motivational thinking — it's structural analysis of what's physically possible versus what's conventionally assumed. The difference between those two things is where most strategic value lives. For a facilitated first-principles session with your leadership team, visit eterdis.com or book a conversation at eterdis.com/contact.