The four metrics that make data estate efficiency measurable.

The four metrics and where each number comes from.

• Percent of the estate unaccessed beyond a defined window, drawn from access-time metadata and age buckets in the index.
• Cost per terabyte by tier, measured against actual usage with a per-GB cost model applied to on-disk data.
• Mean time to data access, from request to grant, taken from access request and permission data.
• Percent of the estate classified by sensitivity, from auto-classification tags across the index.

Where the efficiency actually leaks.

• Redundant, obsolete, and trivial data. Duplicates, orphaned projects, and cold files sitting on tier-1 storage accumulate by default, since nobody buys more storage on purpose. Rule-based cleanup and duplicate detection run continuously with dry-run safeguards, so recovery stays ongoing instead of becoming a project. One data estate reclaimed 30% of tier-1 storage on the first pass.
• Access and permission delays. Every stalled access request is an efficiency cost, not an IT service cost, and it generates ticket volume nobody counts. Permission-aware search and role-based access let the right people reach the right data without a queue in the middle.
• Governance and compliance exposure. Risk that never becomes a number never gets prioritized, which is where most GDPR and CCPA exposure lives today: as an untracked line on a risk register. Auto-tagging by sensitivity turns that coverage into a figure you can report. For the estate Paul and Olivier reviewed, audit prep dropped from three weeks to two days.