amazon-product-research-agent-skills

Amazon Unit Economics and Cash Flow

Purpose and scope

Produce an auditable financial model for one product and marketplace. Separate known costs from assumptions, calculate a base case, and show how the decision changes under plausible scenarios.

Implementation Basis and Dependencies

The included zero-third-party-dependency calculator applies explicit per-unit and first-order formulas. It is a planning model, not accounting, tax, or investment advice.

Dependency	Required	Validation	Fallback
Python 3.10+	Optional	Run python scripts/unit_economics.py --help	Calculate with the documented formulas and label manual work
Price and cost assumptions	Yes	Values are numeric, non-negative, and currency-consistent	Return the missing-assumption list
Marketplace fee basis	Yes for Amazon margin	User-supplied current fee or documented estimate	Do not infer a current fee
Demand and payment timing	Conditional for cash flow	Units, lead time, deposit, balance, and replenishment timing are explicit	Limit output to per-unit economics

Inputs

Provide JSON with the fields documented by scripts/unit_economics.py --example. Required fields are selling price, referral fee, fulfillment fee, landed product cost, return rate, return loss per returned unit, advertising ACOS, and other variable cost. First-order cash analysis additionally needs order units, tooling, setup costs, deposit rate, and working-capital buffer.

Use one currency per run. Percentages are decimals from 0 to 1. Record the source and date of marketplace fees separately in the surrounding evidence bundle. Prices, quotes, and internal margins may be sensitive.

Outputs

The calculator returns JSON containing revenue after returns, Amazon fees, expected return loss, advertising cost, contribution profit and margin, break-even ACOS, first-order cash requirement, and scenario results. It writes to stdout or an explicitly selected file and performs no network or account actions.

Success requires that inputs pass validation, formulas reconcile, units and currency are stated, and the output distinguishes per-unit profitability from inventory cash. Invalid input returns a non-zero exit code without partial financial claims.

Workflow

1. Record marketplace, currency, fee source, quote date, and scenario name.
2. Separate observed costs, calculated fees, modeled rates, and assumptions.
3. Run the base case using scripts/unit_economics.py.
4. Test at least a downside return-rate case and a downside advertising-cost case.
5. Compare contribution margin with the user’s hurdle rate.
6. Calculate first-order cash separately from accounting profit.
7. Identify the assumptions that can reverse the decision and assign a validation action.
8. Return INSUFFICIENT_EVIDENCE when a critical fee or cost is missing.

Example

python scripts/unit_economics.py --example > assumptions.json
python scripts/unit_economics.py assumptions.json --pretty

Error handling and stopping conditions

• Negative price or cost, percentages outside 0..1, or inconsistent currency: fail validation.
• Current Amazon fees are unavailable: request them instead of substituting memory.
• Freight, duty, returns, discounts, or packaging are omitted: show the omission as a decision risk.
• Demand timing is absent: do not claim a cash-flow result.
• Contribution margin is positive but first-order cash exceeds the limit: report the distinction.

Validation and definition of done

• Base, downside-return, and downside-advertising scenarios reconcile.
• Break-even ACOS is never presented as the recommended target ACOS.
• Every financial recommendation names the fee date and critical assumptions.
• python -m unittest discover -s tests -v passes for the bundled calculator.

This site is open source. Improve this page.