[§1] Reference
An inspectable model. A documented judgment
The formulas, assumptions, evidence requirements, and limitations behind Sparrow’s valuation workbench and research notes.
Methodology v1.0 · 26 October 2026. The public workbench is a transparent arithmetic model. A review of an actual dataset requires additional evidence and an agreed scope.
How the workbench computes a scenario
Define a unit first: one record, one hour of source audio, one document, or another consistently measured unit. Enter all prices in USD per that same unit. Compare the same dataset and transaction scope; a sale and a term-limited license convey different rights.
- Usable units = total volume × usable share / 100. Fractional units are permitted where the underlying unit can be divided. Deduplication and exclusions should be reflected in this share.
- Gross proceeds = usable units × entered unit price. There is one separate price for each structure: non-exclusive license, exclusive license, and outright sale. No exclusivity multiplier is inferred.
- Commercial deduction = gross proceeds × deduction percentage / 100. This is one combined reduction; do not enter the same expense here and in costs.
- Base net proceeds = gross proceeds − commercial deduction − preparation and delivery costs. The same costs and deduction percentage apply to all three structures for this comparison. Negative results are shown, not clipped to zero.
- Low / high: multiply each unit price by (1 − variation / 100) or (1 + variation / 100), then repeat steps 2–4. These are specified scenarios, not statistical confidence intervals or estimates of likelihood.
The display rounds USD to two decimals. Calculations use unrounded JavaScript numbers. Very large inputs can lose sub-cent precision. Inputs are bounded to one billion units, one million USD per unit, one trillion USD costs, and 0–100 percent for percentage inputs. These are interface limits, not plausible market boundaries.
Illustrative preset
The optional Illustrative preset uses 100,000 total units, 80% usable share, USD 0.12 / 0.30 / 0.50 per unit, USD 2,000 costs, 10% deductions, and ±25% unit-price variation. These numbers were invented to demonstrate the calculation. They are not reported deals, suggested prices, or market benchmarks.
The resulting illustrative base net proceeds are USD 6,640, USD 19,600, and USD 34,000 respectively. Editing the preset retains the Illustrative label and produces a user-edited scenario; it does not validate the inputs.
Sensitivity analysis
The table uses the non-exclusive base case. It separately raises unit price, total volume, or preparation costs by 10%, and separately raises usable share or commercial deductions by 10 percentage points (capped at 100%). Each test changes only one input. Results are sorted by absolute change in net proceeds. Relative changes are capped at the input limits. A zero starting value stays zero under a relative change. This is a local sensitivity test, not a global ranking of economic importance.
What is outside the calculation
The model assumes one transaction per structure. It does not assume repeat buyers, renewal revenue, payment timing, probability of completion, tax, a discount rate, or an exclusivity opportunity cost. It does not establish ownership, permission to license, demand, or accounting fair value. Recalculate with a different cost assumption if the cost of a sale differs from a license. A separate cash-flow model is needed to compare long-term strategies.
From arithmetic to evidence
For an actual review, record the rights holder and permitted uses, provenance, validation results, usable volume, language and domain coverage, refresh obligations, and transaction duration. Every proposed unit price needs a dated source or an explicitly marked judgment. Unknown information stays unknown; it is not filled with an invented market average.
Prefer an underlying contract or regulatory filing when available. A press-release headline rarely discloses enough to infer a per-unit price. Record the exact rights, term, service obligations, volume, and currency before treating another transaction as a comparable. No proprietary pricing database or live deal feed is connected to this site.
The original research notes describe a qualitative driver framework. Their references to scores, weightings, market adjustments, or a “most likely” case are analytical concepts, not outputs supported by a calibrated public model. Neither the scorecard nor the workbench establishes a probability distribution or independently measured market value.
Readiness scorecard
The readiness scorecard asks eight questions. Each answer receives 0, 1, or 2 points. The total is divided by 16 and multiplied by 100, rounded to a whole number. Each driver is also shown on a 0–10 scale (points × 5). All weights are equal and heuristic. It measures self-reported readiness; it never produces a monetary estimate. No answer verifies rights or regulatory compliance.
Review and publication
Read the editorial policy for how AI-assisted drafting, source checking, human review, and corrections are handled. A review request sends only the information entered in the form, plus calculator assumptions when the optional sharing checkbox is selected.
Working glossary
- Usable volume: the measured part of the asset that remains after exclusions and quality checks.
- Unit price: the entered price for one defined unit under a particular transaction scope.
- Non-exclusive license: permission that leaves room for other licenses, subject to the actual agreement.
- Exclusive license: permission with an exclusivity restriction defined by field, territory, use, or duration.
- Sale: a transfer of the specified asset and rights; the legal scope must be established separately.
- Net proceeds: modeled gross proceeds less the entered deductions and costs; not accounting profit.
- Sensitivity: the change in a result when one assumption changes.