Dates

Extraction of contract dates, term durations, and calculated end dates

Structured data fields are by definition an abstraction of legal language. When in doubt, use citations, the AI chat, and the actual contract text to validate the structured data against your own interpretation.

Overview

The Dates extraction captures explicit dates from the contract, identifies term durations, and calculates when rights periods end. This multi-phase process transforms complex contractual timing into structured, actionable data.

Extraction Focus

This extraction is designed to capture dates and durations that affect when music exploitation rights begin and end. Understanding what is and isn’t in scope helps you interpret the extracted data correctly.

The extraction focuses on the primary exploitation of musical IP (compositions, masters, recordings) by the parties to the agreement. Ancillary rights, third-party rights, and administrative processes are not extracted as calculable terms.

What Gets Extracted

Explicit Dates

Specific dates stated in the contract:

Date Type	Description	Notes
Agreement Date	When the contract was made/dated	Only ONE per contract
Effective Date	When terms take effect	Only with explicit “effective as of” language
Signature Date	When parties signed	Can be per-signatory or general
Start Date	When a term begins	Must reference a specific date value
End Date	When a term ends	Only for rights or agreement terms
Release Date	When a work was/will be released	Must be explicitly stated with a date

Start and end dates are only extracted when explicitly stated in the contract. Dates that would be inferred from other dates (e.g., “commences on signature” when no signature date is provided) are not extracted.

Term Durations

Periods expressed as durations rather than specific dates:

Term Type	Description
Rights	Period during which exploitation rights are granted (e.g., “10 years from delivery”)
Agreement	Overall contract term, when separate from rights period
Administration	Period of administration control (publishing/neighbouring rights)
Collection	Post-rights period for collecting royalties (publishing agreements)
Retention	Period after term ends before rights revert
Extension	Optional periods that extend base terms (each option is separate)
Notice	Required notice period to terminate and revert rights

Each term includes:

• Duration: Either a fixed value and unit (e.g., “5 years”, “18 months”, “perpetual”) or a range with a minimum (floor) and optional maximum (ceiling) — e.g., “at least 2 years, up to 5 years”
• Description: Context from the contract
• Condition: For variable-bound terms, the trigger that determines where in the range the term actually lands (e.g., “until recoupment”, “until fulfilment of Minimum Commitment”). For fixed terms, only set when start/end depends on an uncertain event (recoupment, option exercise).
• Identifier: Specific catalog or work name (not generic period names)

Duration Structure

Term durations can be either a single exact value or a variable range. The display reflects which kind of duration was extracted:

Contract Language	How it’s displayed
”5 years”	5 years
”perpetual”, “in perpetuity”, “life of copyright”	Perpetual
”minimum 12 months, longstop 3 years”	12 months — 3 years
”at least 1 year, until recoupment”	1 year+
”90 days notice”	90 days

• Fixed durations are used for exact, deterministic terms.
• Floor / ceiling durations are used when the contract defines a single term whose actual length is variable — a guaranteed minimum (floor), an absolute maximum (ceiling), or both. When only a floor is present, the term is displayed with a trailing + to signal the open upper bound.

A variable-bound term (floor/ceiling) represents one period with a flexible length, not multiple periods. Optional extensions that a party can choose to exercise are extracted as separate extension terms instead.

Calculated End Dates

The system calculates when rights periods end based on extracted dates and durations:

End Type	Description
Fixed	Single definitive end date (no variations/options)
Earliest	Soonest the term could end (no extensions exercised)
Latest	Latest the term could end (all extensions exercised)

Each calculated date includes:

• Date: The calculated end date (YYYY-MM-DD)
• Type: rights, agreement, administration, collection, or retention
• Estimated: True if using proxy dates (e.g., release date instead of delivery date)
• Linked assets: Recording and/or composition IDs this date applies to

Perpetual Terms

Terms with “perpetual”, “in perpetuity”, “life of copyright”, or similar language have no calculated end date. These are captured as term durations but do not produce end date calculations.

How Calculated End Dates Stay Up to Date

Calculated end dates are automatically recalculated when the data they depend on changes. The system combines the contract’s terms and durations with asset-specific dates (like release dates) to compute end dates — so when either side changes, the results update.

Change	What Happens
Contract first processed	End dates calculated from extracted dates, terms, and linked assets
Recordings or compositions linked	End dates recalculated using the updated asset set
Commitment links changed	End dates recalculated with the updated commitment-to-asset mapping
Missing dates added	End dates recalculated — estimated dates replaced with concrete ones

Recalculation is automatic and does not require a manual trigger.

Recalculation replaces all calculated end dates for the contract, including any manual edits. Link all the correct recordings and compositions first, then make manual adjustments — otherwise your edits will be overwritten the next time recalculation runs.

Commitment-Scoped End Dates

When a contract has commitments with linked assets, each commitment can produce its own set of end dates for its specific assets.

For example, a contract might grant rights to Album A (recordings 1—5, 5-year term) and Album B (recordings 6—10, 7-year term). Each commitment produces different end dates for its recordings, so rights revert at different times depending on which commitment an asset belongs to.

If a new commitment is added with linked assets but no end dates have been calculated for it yet, the system automatically generates the missing calculations during the next recalculation.

Missing Dates

When a calculation requires a date that isn’t available, it’s listed as missing:

Field	Description
Name	Short name for the missing date (e.g., “Delivery Date”)
Description	Why it’s needed and what calculation it affects

Missing dates only appear when the contract explicitly references a specific date type that isn’t provided. The system does not speculate about what dates “might” be needed.

Reading Dates Data

Understanding Calculated End Dates

When interpreting calculated end dates:

1. Fixed dates mean there are no optional extensions - this is the definitive end
2. Earliest/Latest pairs mean extensions exist - earliest assumes no options exercised, latest assumes all options exercised
3. Estimated dates use proxy dates (e.g., release date when delivery date is unavailable) - verify against actual dates when available
4. Per-asset dates (showing linked recordings/compositions) mean each asset has its own timeline based on its release date

A Multiple dates badge appears next to a recording or composition in the linked assets list when that asset is tied to more than one calculated end date of the same type (e.g., two different rights end dates). Use it as a signal to expand each row and compare the dates side-by-side.

Why an Asset Can Have Multiple End Dates

Multiple end dates for the same asset can occur for several reasons:

• Different date types — Rights end in 2028, administration ends in 2030. This is normal; different rights revert at different times.
• Earliest and Latest pairs — Extensions exist. Earliest assumes no options exercised, Latest assumes all options exercised. These are complementary, not conflicting.
• Different commitment scopes — The asset appears in multiple commitments with different terms (e.g., an initial period and an option period).
• Unresolved identifiers — The contract specifies different terms for different catalog segments (e.g., “Album A” vs. “Album B”), but the system could not determine which segment a particular asset belongs to. In this case the asset is assigned to all matching end dates so it isn’t lost. The Multiple dates badge alerts you to review and remove the asset from the end dates that don’t apply.

The Timeline and Categorized Dates views split these into separate entries so you can review each one individually.

When you see the Multiple dates badge, expand the rows to compare. If the dates come from different identifiers (catalog segments), it usually means you need to remove the asset from the incorrect end date — the system included it in all of them to make sure it wasn’t missed.

Date Hierarchy

The extraction follows a clear hierarchy:

Contract Document
│
├── Explicit Dates (stated in contract)
│   ├── Agreement Date, Effective Date
│   ├── Signature Dates
│   ├── Start/End Dates (explicit)
│   └── Release Dates
│
├── Term Durations (periods stated in contract)
│   ├── Rights, Agreement, Administration
│   ├── Collection, Retention
│   ├── Extensions (each option separate)
│   └── Notice periods
│
├── Calculated End Dates (computed)
│   ├── Fixed OR Earliest/Latest
│   └── Linked to assets
│
└── Missing Dates (needed for calculations)

Editing Dates

Click the edit icon to:

• Add missing dates (delivery dates, release dates, etc.)
• Correct date values
• Modify term durations
• Update term descriptions and conditions
• Add, edit, or remove calculated end dates and the assets they apply to

Adding missing dates triggers recalculation of end dates, giving you accurate reversion timelines.

Editing Term Durations

When adding or editing a term duration, pick a Duration Type before entering the value:

1. Fixed duration

   A single, definite period — e.g., “5 years”, “18 months”. Enter a Value and pick a Unit (year, month, week, day, or perpetual).

   Select perpetual as the unit when the contract uses “in perpetuity”, “life of copyright”, or similar language. The value field is automatically zeroed out and disabled, and these terms produce no calculated end date.

2. Range (floor / ceiling)

   A duration expressed as a minimum and an optional maximum — useful when the contract sets a minimum term that can extend up to a cap (e.g., “at least 2 years, up to 5 years”).

   • Minimum (floor) — required. Sets the shortest the term can run.
   • Maximum (ceiling) — optional. Click Add ceiling to set a cap; click Remove ceiling to drop it.

   The duration summary at the bottom of the dialog shows the range — 2 years — 5 years when a ceiling is set, or 2 years+ when only a floor is set.

Editing Calculated End Dates

Adding or editing a calculated end date (Fixed, Earliest, or Latest) opens a three-step flow:

1. Details

   Set the date itself and its metadata:

   • Date — the calculated end date (YYYY-MM-DD)
   • Type — one of rights, agreement, administration, collection, or retention
   • Identifier — optional catalog or work name this date applies to
   • Description — context about how the date was calculated
2. Link Recordings & Compositions

   Pick which assets this calculated end date applies to. Only recordings and compositions already linked to the contract can be selected — link them in the Assets dialog first if they’re missing here.

   Each linked recording or composition opens a link to its catalog page so you can verify you’re picking the right asset. Use the X button to unlink an asset from this date without removing it from the contract.

3. Preview

   Review the date, its linked assets, and the change summary before saving. The preview shows a count of linked recordings and compositions so you can confirm the scope at a glance.

Linking specific recordings and compositions to a calculated end date is what powers per-asset reversion timelines. If a calc date applies to the whole contract, leave both lists empty; if it applies only to specific tracks or works, link them here so Categorized Dates and the Timeline can split the reversion records correctly.

Common Scenarios

Adding Assets After Initial Extraction

A common workflow is uploading a contract before all its assets are linked:

1. Contract uploaded — The contract says “5 years from delivery of each recording”. The system extracts the term but produces no calculated end dates yet because no recordings are linked.
2. Recordings linked — You add recordings to the contract. Recalculation runs automatically and produces end dates for each recording.
3. Missing delivery dates — If the recordings have release dates but no delivery dates, the release date is used as a proxy and the calculated dates are flagged as Estimated.
4. Delivery date added later — You add the actual delivery date to the contract’s explicit dates. Recalculation replaces the estimated dates with concrete ones based on the real delivery date.

Calculated end dates only appear once assets are linked — the contract’s terms alone aren’t enough to produce a date when the calculation depends on per-asset dates like release or delivery dates.

Mixing Perpetual and Fixed Terms

A contract can combine perpetual and fixed terms for different rights or assets. For example, publishing rights might be granted in perpetuity while administration rights have a 10-year term. The perpetual rights produce no calculated end date, while the fixed-term rights do. Both are captured accurately — check the Timeline to see only the dates that have a calculable expiration.

See Also

Categorized Dates

View reversion dates organized by unique asset, date, and territory combinations. Optimized for tracking reversions, timeline views, and exports.