Cost Benefit

Cost-Benefit

Tally the full costs and benefits of an action — including opportunity cost, hidden costs, and to-whom-they-accrue — before judging it worthwhile.

Why

Untrained tallies count visible, near-term, own-side effects and systematically miss opportunity cost, ongoing maintenance burden, and externalities — which flatters taking action over not taking it, every time.

Use when / Don't use when

• Use when: justifying or challenging any investment of money, time, or attention.
• Don't use when: stakes are genuinely trivial, or the decision is so dominated by a single factor that a full ledger adds nothing.

Inputs → Outputs

• Inputs: a proposed action.
• Outputs: a ledger of costs and benefits with timing, bearer, and confidence per entry, plus a net judgment flagging the dominant uncertainties.

Principles

• The opportunity cost — the best foregone alternative — is a real cost, and it's almost always missing from the first draft of any ledger.
• Count ongoing costs (maintenance, attention, added complexity), not just the upfront acquisition cost.
• Note who bears each cost specifically — a net-positive action can still be a transfer of cost onto someone who was never consulted.
• Sunk costs are not costs, and shouldn't appear in the ledger at all.

Procedure

1. State the action and the actual comparison baseline — usually the best alternative use of the same resources, not simply "do nothing."
2. List benefits with their timing and confidence.
3. List costs: direct, ongoing, attention/complexity, and risk-bearing.
4. Add the opportunity cost explicitly as its own line item.
5. Mark who bears each entry, and flag any asymmetries — benefits accruing to one party, costs to another.
6. Net it out, and state which one or two uncertain entries actually dominate the conclusion — that's the residual uncertainty worth resolving further.

Common mistakes

• Using "do nothing" as the baseline when a live, better alternative actually exists.
• Ignoring the long tail of maintenance costs that follow an initial investment.
• Counting sunk costs as if they belonged in the ledger.
• Treating hard-to-quantify entries as if they were zero, rather than estimating them explicitly.

Examples

• Whether to build an internal tool versus buying or doing without.
• Whether a recurring meeting series earns back its collective hours across attendees.
• Whether a regulation's compliance burden is proportionate to the protection it provides; whether to adopt a new dependency.

Related

• [[expected-value]] — the probabilistic version, used when entries are genuinely uncertain rather than point estimates.
• [[second-order-scan]] — often surfaces the hidden entries this ledger would otherwise miss.
• [[prioritization-triage]] — the natural next step once several cost-benefit ledgers need to be compared against each other.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: skyf0xx
• Source: skyf0xx/better-thinking
• License: MIT
• Homepage: https://zenbin.org/p/better-thinking

Install and usage instructions live in the source repository linked above.