Reversibility-Stakes Assessment

Assess decisions by magnitude and reversibility—most feel high-stakes but aren't

Gibson Biddle
The art of product strategy and prioritization

Reversibility-Stakes Assessment

"As product managers, we feel like every decision we make is high stakes. It's good occasionally to think about, hey, what's the magnitude? And then, is it reversible?" - Gibson Biddle

What It Is

Reversibility-Stakes Assessment is a framework for evaluating how much time and energy to invest in any decision. It combines two dimensions:

  1. Magnitude: How big is the potential impact relative to the overall business?
  2. Reversibility: Can we undo this decision if it doesn't work out?

The framework helps product managers (and anyone else) recognize that most decisions feel high-stakes but aren't. This awareness prevents analysis paralysis and allows faster decision-making where appropriate.

Amazon famously calls this the "one-way door vs. two-way door" distinction.

How It Works

The Two Dimensions

Magnitude Assessment:

  • What's the potential downside in absolute terms?
  • What's the downside relative to company scale?
  • Is this an existential risk or a recoverable mistake?

Reversibility Assessment:

  • Can we undo this decision?
  • How hard/expensive would reversal be?
  • Is this a two-way door (reversible) or one-way door (irreversible)?

The Netflix Auto-Cancel Example

Netflix noticed 0.5% of members weren't using the service—they'd signed up a year ago and forgotten they were paying. The decision: should Netflix auto-cancel these inactive members?

Magnitude Analysis:

  • Cost: ~$100M in revenue (0.5% × ~$20B revenue)
  • Context: Company doing $30B in revenue
  • Relative magnitude: ~0.3% of revenue

Reversibility Analysis:

  • Can they stop auto-canceling later? Yes
  • Can they re-acquire those customers? Possibly, if customers want to return
  • Is this reversible? Yes—it's a two-way door

Decision: Despite the $100M cost, this was actually a low-stakes decision because:

  1. Magnitude was small relative to company size
  2. The practice could be reversed if needed
  3. The brand benefit (delight + hard-to-copy advantage) was worth exploring

The Marriage Analogy

Biddle uses marriage as a contrast:

"I got married 30 years ago. I was anxious about getting married. My friend said, 'Gib, if it doesn't work out you can always get divorced.' So they're saying it's reversible. Now I've been married 30 years, so it's worked out."

His point: Even "irreversible" decisions often have escape hatches. But some decisions (marriage, having kids) are fundamentally different in character—they're one-way doors where reversal, while possible, comes with significant consequences.

How to Apply It

Quick Assessment Matrix

Low Magnitude High Magnitude
Reversible Just do it Move fast, monitor closely
Irreversible Careful consideration High-stakes analysis

Step-by-Step Process

  1. Estimate magnitude:

    • What's the worst-case scenario?
    • What percentage of revenue/users/resources does this affect?
    • Compare to recent decisions you've made

  2. Assess reversibility:

    • Can we undo this?
    • What would reversal cost (time, money, relationships)?
    • How long until we know if it's working?

  3. Calibrate your investment:

    • Low magnitude + reversible: Decide quickly, iterate
    • High magnitude + irreversible: Take time, get input, stress-test

  4. Avoid the PM trap:

    • Notice if you're treating low-stakes decisions as high-stakes
    • Ask: "Am I overthinking this?"

When to Use It

  • Feature launches: Is this reversible? How big is the blast radius?
  • Pricing changes: Can we roll back? What's the revenue impact?
  • Organizational decisions: Can we reorganize if this doesn't work?
  • Career decisions: Is this genuinely irreversible or just feels that way?
  • Time allocation: Am I spending high-stakes energy on low-stakes decisions?

Examples by Category

Genuinely High-Stakes (One-Way Doors)

  • Major acquisitions
  • Platform migrations that break backward compatibility
  • Shutting down a product line
  • Firing key executives
  • Taking on major debt

Feels High-Stakes But Isn't (Two-Way Doors)

  • Most feature launches (can be rolled back)
  • Pricing experiments (can be adjusted)
  • Process changes (can be revised)
  • Hiring decisions (not great to reverse, but possible)
  • Most A/B tests (that's literally the point)

The Meta-Insight

"As product managers, we feel like every decision we make is high stakes."

The framework's real value is calibrating your emotional response to decisions. Most product decisions:

  • Affect a small percentage of users
  • Can be reversed if they don't work
  • Won't make or break the company

Save your high-stakes energy for genuinely high-stakes moments.

Source

  • Guest: Gibson Biddle
  • Episode: "The art of product strategy and prioritization"
  • Key Discussion: (27:50-29:04) - Discussion of high vs. low stakes decisions using Netflix auto-cancel example
  • YouTube: Watch on YouTube

Related Frameworks

Back to Decision Making