Zero Based Budget Planner

Zero-Based Budget Planner

When to use

Use whenever spend needs to be justified from a zero base rather than carried forward from last year, or when a budget cut is coming and the instinct is an across-the-board percentage reduction — a method that cuts high-value and low-value spend by the same amount, protecting waste exactly as much as it protects what actually matters.

What it does

Scores every discretionary spend item as a decision package with a cost and a value score, computes value per dollar via a bundled calculator, ranks all packages by that ratio, and shows precisely which packages a stated total budget actually funds versus cuts — separating mandatory spend (compliance, core operations) that gets funded regardless of ranking from discretionary spend that competes on value.

Method

1. Separate mandatory from discretionary spend up front. Mandatory means the business cannot legally or operationally function without it (compliance, core infrastructure, contractual obligations) — these get funded first and are excluded from the ranking competition, since ranking them against a marketing campaign would be a category error.
2. Break discretionary spend into decision packages — a specific initiative or program with a cost and a stated business value, not a department-level lump sum. A department budget line is too coarse to rank; the individual programs inside it are what actually compete for funding.
3. Score each package's value (1-10) based on its actual business impact — revenue contribution, risk reduction, strategic importance — with the reasoning stated, not just a number pulled from department politics.
4. Run the bundled calculator (scripts/zbb_ranker.py) to compute value per dollar for every discretionary package, rank them, and show exactly which packages fall inside the available budget and which get cut.
5. Read the cut list as a set of explicit trade-offs, not a byproduct. If a cut package is politically important despite ranking low on value per dollar, that's now a visible, named decision to override the ranking, rather than an invisible casualty of a blanket percentage cut.
6. Check for packages that are individually low-value but collectively necessary (a set of small tools that together support one critical workflow) — the value score should reflect the package's real function, not just its individual visibility.
7. Publish the ranking, not just the final funded list. Showing why each package was funded or cut is what makes a zero-based budget defensible when someone asks "why did we cut X but keep Y."

Inputs

• Total budget available
• List of decision packages with cost, value score and rationale, and mandatory flag where applicable
• Config saved as JSON matching the format documented at the top of scripts/zbb_ranker.py

Output format

Mandatory spend total and remaining discretionary budget; ranked table of discretionary packages by value per dollar with funded/cut status; explicit cut list with a note that political overrides of the ranking are visible trade-offs, not silent losses.

Example

A $3M budget with $1.2M in mandatory core platform spend leaves $1.8M discretionary. Ranking four discretionary packages by value per dollar funds customer success tooling, market expansion, and a marketing campaign, while an office renovation ranks last and gets cut. If leadership still wants the renovation funded, that's now an explicit override of a stated ranking, not a spending decision made by default because nobody compared it against the alternatives.

Common pitfalls

• Cutting every department by the same percentage instead of ranking individual spend items by actual value.
• Scoring value based on who's loudest in the budget meeting rather than a stated, defensible rationale.
• Treating mandatory and discretionary spend as if they compete on the same ranking, when they answer different questions.

Source & license

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

• Author: Natan-Mohart
• Source: Natan-Mohart/24-finance-skills-for-claude
• License: MIT

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