First Principles or Best Practice? A 60-Second Decision Test

Here’s the question nobody talks about: when should you NOT use first-principles thinking?

First-principles thinking is powerful when it’s needed. But it’s also expensive — it takes time, energy, and cognitive effort to decompose a problem to fundamentals. And for many problems, a proven framework or best practice is the better choice.

The trick is knowing which tool fits the situation. This article gives you a 60-second test to decide.

The cost of always reasoning from scratch

Imagine you’re setting up a new development team. You could reason from first principles about how humans collaborate on code — examining every assumption about version control, code review, deployment, and communication.

Or you could adopt a well-tested workflow (git flow, trunk-based development, whatever fits your scale) and start shipping.

The first-principles approach might eventually produce a novel insight. But for a standard problem with proven solutions, the cost of re-deriving everything from scratch far exceeds the benefit.

As one Hacker News commenter put it: “First principles thinking is expensive and doesn’t scale. You’re welcome to make more bespoke blocks, but iterating from first principles every time is just not necessary when we can rely on the common wisdom gained by using more common toolkits.”

That’s a valid objection. The answer isn’t to abandon first principles — it’s to be selective about when you use them.

The 60-second decision test

Ask these four questions about your problem. If you answer “yes” to two or more, use first-principles thinking. Otherwise, start with a proven approach.

Question 1: Is the problem stuck?

Have you or your team been working on this for a while without meaningful progress? Are you cycling through variations of the same approach? When a problem resists standard solutions, it’s usually because the framing itself contains an unexamined assumption.

Examples:

• A product team has tried three different onboarding flows but activation rates won’t budge
• A pricing strategy keeps getting tweaked but revenue stays flat
• A process has been “optimized” multiple times but remains slow

If the problem is stuck, first principles can break the loop by questioning the frame, not just the solution.

Question 2: Are the stakes high enough to justify the cost?

First-principles analysis takes real time and effort. Is the potential payoff worth it?

Worth it:

• Choosing your company’s core product direction
• A decision that locks you into a multi-year commitment
• A problem where being wrong costs more than being slow

Not worth it:

• Picking a project management tool
• Deciding on a meeting format
• Choosing between two similar vendors

For routine, reversible, low-consequence decisions — just pick a proven option and move on.

Question 3: Is the situation genuinely novel?

Has someone already solved this problem well? If hundreds of companies have faced the same challenge and converged on a similar solution, that convergence is information. It doesn’t prove the solution is optimal, but it shifts the burden of proof.

Novel situations where first principles help:

• You’re entering a market that didn’t exist two years ago
• The constraints of your situation are materially different from standard advice
• The “standard answer” was designed for a context that doesn’t match yours

Not novel:

• Setting up accounting for a small business
• Implementing user authentication
• Writing a job description

Question 4: Do you suspect the “best practice” rests on outdated assumptions?

This is the strongest trigger. If you look at the conventional approach and can articulate which specific assumption might be wrong, first-principles thinking becomes valuable.

Signs the best practice may be stale:

• The practice was established before a key technology or market shift
• Everyone follows it but nobody can explain why it works
• The practice optimizes for a constraint that no longer applies
• Experts disagree about whether it’s still the right approach

If you can’t identify a specific questionable assumption, the best practice is probably fine.

The flowchart

Is the problem stuck?
  YES ─┐
  NO   │  Are the stakes high?
       │    YES ─┐
       │    NO   │  Is it genuinely novel?
       │         │    YES ─┐
       │         │    NO   │  Do you suspect stale assumptions?
       │         │         │    YES ─ USE FIRST PRINCIPLES
       │         │         │    NO  ─ USE BEST PRACTICE
       │         │         │
       │         │    ─────┘
       │    ─────┘
  ─────┘

Score: 2+ YES answers → First principles
       0-1 YES answers → Best practice (and revisit if it stops working)

A worked example

Problem: Your sales team’s close rate has dropped from 30% to 18% over six months.

Let’s run the test:

1. Is it stuck? Yes — the team has tried new scripts, additional training, and incentive changes with no improvement.
2. Are the stakes high? Yes — this directly affects revenue and the team may need to be restructured.
3. Is it novel? Partially — sales processes are well-studied, but the drop happened after a market shift to remote buying.
4. Do you suspect stale assumptions? Yes — the entire sales process was designed for in-person demos.

Score: 3-4 YES answers. Use first principles.

The team would then decompose the problem: What does the buyer actually need to make a decision? Which parts of the demo are essential information vs. in-person theater? What assumptions about “how we sell” are inherited from a pre-remote world?

This kind of decomposition might reveal that the demo itself isn’t the problem — the problem is that the qualification process sends the wrong prospects to the demo. Or that remote buyers need different proof points than in-person ones. You can’t find these insights by optimizing the existing script.

When to switch modes

Even after choosing an approach, stay flexible:

• Started with best practice but it’s not working? Switch to first principles. The failure itself is useful data — it tells you which assumption in the best practice doesn’t hold for your situation.

• Started with first principles but the decomposition isn’t revealing anything new? Switch to best practice. If your fundamentals look the same as everyone else’s, the problem isn’t in the frame — it’s in the execution.

• Built something from first principles and it’s working? Document what you learned and turn it into your own best practice. First principles is for discovery; repeatable processes are for scale.

The key distinction

First-principles thinking and best practices aren’t opposites. They’re different tools for different situations:

The goal isn’t to always reason from scratch. It’s to know when reasoning from scratch is worth it — and when the smartest move is to use what already works.

Want the full process for when you do choose first principles? The Problem Deconstruction Worksheet walks you through 8 steps — from stating the real problem to selecting your cheapest test. Free, printable, no signup required.