The arithmetic mean elapsed time from a physical item's formal intake into the digitization workflow until its digital surrogate is processed, quality-checked, metadata-applied and made available for the intended access channel. Measurement requires explicit start and end points, stated inclusion rules for queueing and post‑processing, and a defined aggregation method (mean, trimmed mean, or median).

Principle

As a throughput metric, the mean time to digitization summarizes end-to-end process capacity but is sensitive to batching, outliers and the chosen aggregation rule; it therefore reflects operational workflow and resource allocation rather than an invariant per-item processing time.

Demonstration

Illustrative scenario → A repository records start times when items are received and end times when digital objects pass final QA and are published. For 100 items with total elapsed times summing 5,000 hours, Mean Time to Digitization = 50 hours per item. Reporting specifies whether times include queue wait, conservation work, or deferred embargo processing.

Misapplication

Treating the mean as the definitive per-item processing duration without documenting whether queueing, parallel processing, or excluded subwork (e.g., conservation) were omitted; this confuses throughput with average active labor and underestimates variability from outliers.

Consequence

Used for capacity planning, staffing, SLA definition and project costing; an inaccurately or inconsistently measured mean can mislead resource commitments, produce unrealistic deadlines, or conceal backlog problems.

Reversal

When workflows produce heavy-tailed time distributions (frequent long outliers) or when user‑facing SLAs require predictable delivery, median or percentile metrics (e.g., 90th percentile) or stratified means (by item type) may be more informative than the arithmetic mean.

Boundary

Clearly within: defined batch digitization of physical items where intake and availability events are recorded. Boundary case: items that undergo unpredictable conservation or vendor outsourcing—measurement must state inclusion. Clearly outside: measures of born‑digital ingest or downstream access latency after publication.

Semantic Tension

Throughput (speed) ↔ Fidelity (quality and preservation): reducing mean time often pressures quality, whereas prioritizing fidelity increases mean time; measurement must make that trade‑off explicit.

Synthesis

Mean Time to Digitization is a process-definition-dependent throughput indicator: useful for comparing workflows and allocating resources only when its start/end definitions, inclusion rules, and aggregation method are explicit and aligned with operational goals.