The Data Definition Card
The data definition card is the full form of a definitions register entry: the ten fields of The Definitions Register extended with the formula, the unit, the instrument class that captures the number, the source row that supplies it, the validation rule it must pass and the lineage cards that consume it, so that an auditor walking a capture point can map it to one card and a planner reading a model can walk back from the model to the instrument. It matters because the register's entry was built to settle what a word means, and a multichannel operation needs beside it the facts about how the number is produced, which is where asynchronous channels fail: the definition of email handling time can be agreed in an afternoon, and the card is where the function discovers that no instrument produces it. The card is part of the Workforce Data Definitions Framework; the register remains the index, the card is the entry in full, and both change only through GP-606.
The card's shape follows the metadata-registry standard, in which a data element is the pairing of a concept, what is being measured, with a representation, the unit and value domain it is measured in, and both are registered under one identifier with a responsible steward,[1] and it follows Deming's rule that a measurement without an operational definition, the procedure that produces it, has no meaning.[2] The practice of publishing a structured sheet beside a dataset, stating how it was collected, what it contains and what it should not be used for, has the same motive in the machine-learning literature.[3]
What the card adds to the register row
The register entry's ten fields are kept unchanged, the Instrument field among them, which continues to name the system, the field and the instrument's version. The card adds the fields an audit and a model need and the register row cannot hold without becoming unreadable.
| Field | Holds | Rule |
|---|---|---|
| Id | The metric code and version (AHT-02 v1), the register's RG number beside it |
Never reused; a changed definition is a new version |
| Name | The metric's name, qualified by channel where the channel changes the definition | Handle time on voice and handling time on an asynchronous channel are two cards, because the unit and the instrument class differ |
| Definition | One plain sentence | The sentence leads; the formula follows |
| Formula | Numerator, denominator and summation stated with their terms named | Every term in the formula is itself a card or a ledger bucket on Agent Time Accounting Model |
| Unit | What one of the number is: seconds per handled contact, minutes per item fulfilled, percent of logged-in time | A number without a unit cannot be rolled up; Workload Units and the Roll-Up to FTE carries the ladder the unit sits on |
| Instrument | Carried from the register: the system, the field and the instrument's version | Unchanged |
| Instrument class | The class of capture mechanism: platform state timer, work-item timer, desktop focus-time instrument, time-and-motion sample, statistical inference | A class, never a product; the methods and their biases are on Email Handle Time Measurement and Chat Concurrency and Handle Time |
| Resolution | What the instrument can distinguish: per contact, per touch, per interval, per cohort | Carried from the register; a comparison below the resolution is noise reported as a finding |
| Source | The ST row on the Source of Truth Register the number is taken from | Exactly one; the card's grade ceiling is the row's source grade |
| Validation rule | The reconciliation test on Reconciliation Tests for Time Accounting the number must pass, and the plausibility range outside which a value is quarantined | The gate on Data Quality Gates for Planning Inputs runs it; a card with no validation rule has no way to fail |
| Lineage | The LN cards on Data Lineage Card for Planning Models that consume the metric | Derived; the list GP-606 reads to find what a change moves |
| Pairing | What the number must be read beside | Carried from the register |
| Owner | The seat that answers questions about the card | A seat, never a name |
| Signatories | The seats that adopted it, jointly | The registers' signatories; a card signed by one seat is a private definition |
| Version and effective date | Which version applies from when | Forward only; previous versions kept |
| Carry rule | Whether the number survives a channel, platform or cohort change | Carried from the register: [A] in the new state until re-measured |
| Change history | Every proposal, review and adoption | Shared in form with the registers |
The register's AHT-01 already states its formula "by channel", and the card does not split a code for every channel. Asynchronous handling time is split out as AHT-02 because the split is not in the formula's terms but in the unit and the instrument class: seconds per handled contact read from a state timer, against minutes per item fulfilled summed over touches read from a work-item timer, and one code carries one unit.
Card, source row and lineage card
Three objects form a triangle, and each cites the other two. The definition card says what the number means and how it is captured; its source field cites one ST row. The ST row says which system the number is taken from; its metric field cites one card. The lineage card says which model consumes the number; its inputs cite both. A change to any corner is traced to the other two in one step, which is the property GP-606 relies on and the reason the three share one change control and one set of signatories rather than three. A card whose source field is empty is a definition without a number, carried at grade Open with an owner and a date, as the register carries ramp; a card whose lineage field is empty is a number nobody uses, which is a finding for GP-605 rather than a defect in the card.
Why the instrument class is the field that matters on an asynchronous channel
On voice the instrument is the telephony platform's state timer, and the definition and the instrument agree without effort because the platform was built to produce the number the definition names. On email and case work there is typically no in-work state: the item has an open time and a close time, the agent's time on it is spread across touches separated by pending periods, and the work is often done while the agent sits in a voice available or auxiliary state, so the time appears in the ledger as idle or as another activity. The register entry's instrument field, "the case tool's work-item record", reads as if an instrument exists. The card's instrument class forces the question: a timer on the touch, a focus-time instrument on the desktop, a sample, or an inference? Each answers to a different source grade and bias, and Production Accounting for Blended Agents gives the contamination the validation rule must test for. The same forcing applies to chat, where the instrument must separate session elapsed time from agent focus time or the card defines a number that doubles under concurrency.
The filled example
One card, synthetic, drafted for a function's register v0 on Thu 16 Oct 2026. Numbers are illustrative and graded [E].
| Field | Value |
|---|---|
| Id | AHT-02 v0 (RG-008)
|
| Name | Handling time, asynchronous channel (email and case) |
| Definition | The agent work time spent on one item from first open to close, summed across touches, excluding the time the item waits in queue or pends between touches |
| Formula | HTitem = Σ touch durationk over the item's touches; elapsed time (close − arrive) held as a second field, never blended |
| Unit | Minutes per item fulfilled |
| Instrument | The case tool's work-item record, touch-timer field, instrument version stated at adoption; until then, the time-and-motion sample's handling-time field |
| Instrument class | Work-item timer, reconciled to a time-and-motion sample; until the timer exists, the sample alone |
| Resolution | Per touch; per item; per interval by close time; per cohort and tenure band |
| Source | ST-007 (v0: the sample, source grade named owner; v1 target: the timer with the sample, system of record)
|
| Validation rule | Test T-05 item-count against time: the ratio of items closed × handling time per item to in-work time for the channel on the ledger within tolerance of 1.0, two-sided as the test states, tolerance SET LOCALLY; a value outside the plausibility range (SET LOCALLY) is quarantined |
| Lineage | LN-004 email requirement and backlog; through it, LN-003 multichannel requirement
|
| Pairing | Read with items closed and backlog age; never beside voice handle time as one blended figure |
| Owner | The definitions owner (a seat) |
| Signatories | The forecasting lead for asynchronous channels and the data and analytics lead, jointly |
| Version · effective | v0 · Thu 16 Oct 2026 (proposed); v1 target at the timer's adoption |
| Carry rule | [A] in any new case tool, channel or cohort state until re-measured |
| Change history | v0 drafted Thu 16 Oct 2026; adoption pending the signatories' review slot |
The card's value is in what it refuses. It refuses elapsed time as handling time, which is the commonest substitution on an asynchronous channel and overstates the requirement by the pending time. It refuses an undeclared instrument, by naming the sample as the standing source until the timer is adopted, so that every figure computed on it is [E] with the sample's range. And it names the test that catches work done inside a voice state, so that the contamination Production Accounting for Blended Agents describes is a validation failure rather than an invisible inflation of availability.
Maturity Model Position
Four scales on this wiki use the word level; the launch page states which is which. On the WFM Labs Maturity Model™ a register entry with an owner is Level 2 work. A card per metric with instrument class, source row, validation rule and lineage, signed jointly and cited by id and version from every model, is the Level 3 form, and the form an agent reads when it asks what a column means. At Level 4 the validation rule's plausibility range is the prior a distributional plan starts from.
See Also
- Workforce Data Definitions Framework — the hub: the ledger, the definitions and the sources every model computes on
- The Definitions Register — the ten-field entry the card extends; the six core definitions
- Source of Truth Register — the ST row the card's source field cites
- Data Lineage Card for Planning Models — the LN cards the lineage field lists
- Data Quality Gates for Planning Inputs — the gate that runs the validation rule
- Reconciliation Tests for Time Accounting — test T-05 and its tolerance
- Agent Time Accounting Model — the ledger buckets the formula's terms land in
- Email Handle Time Measurement — the instrument classes and their biases
- Production Accounting for Blended Agents — the contamination the validation rule tests for
- Definitions and Source-of-Truth Change Control (generic process) — GP-606, the card's change control
References
- ↑ International Organization for Standardization / International Electrotechnical Commission (2023). ISO/IEC 11179-1:2023 Information technology — Metadata registries (MDR) — Part 1: Framework.
- ↑ Deming, W. E. (1986). Out of the Crisis. MIT Center for Advanced Engineering Study, ch. 9 (operational definitions). ISBN 0-911379-01-0.
- ↑ Gebru, T., Morgenstern, J., Vecchione, B., Vaughan, J. W., Wallach, H., Daumé III, H., & Crawford, K. (2021). "Datasheets for Datasets". Communications of the ACM 64(12), 86–92. doi:10.1145/3458723.
