InSpective

Tier 1 · Configuration

Configuration integrity & standards alignment

A Sharedo tenant accretes configuration faster than it sheds it. Fields are added for one matter type and never removed; the same concept ends up stored two ways; forms grow to ask for data no one supplies. This audit measures every attribute field's fill rate over the live matter population, tests the estate against Alterspective's library of 19+ active Sharedo configuration standards, and condenses the result into a health score out of 100 whose indicator values, thresholds, and weights are printed on the report. You get a prioritised, rule-referenced clean-up backlog, not a pile of observations.

T1

1What we examine

Form-field estate & fill ratesCFG-FF

Every attribute field is measured for fill rate across the live matter population. Dead fields (0% populated) and sparse fields (under 5%) are identified and ranked: the signature of forms asking for data no one supplies.

Mixed-representation defectsCFG-OS

The same concept stored two ways at once: free-text values in one field and option-set codes in another, which quietly breaks reporting and rule logic.

Single-field value driftCFG-FF

One field carrying true/false in some rows and 1/0 in others. Reports built on it silently under-count, and nobody notices until a number is challenged.

Duplicate field namesCFG-FN

Logical fields configured twice under keys differing only by case or punctuation (for example reserve-amount versus ReserveAmount), so data splits across both copies and neither is complete.

Standards alignmentCFG-WT · CFG-PH

Configuration is measured against our standards library of 19+ active standards, covering work types (CFG-WT), phases (CFG-PH), key dates (CFG-KD), option sets (CFG-OS), form fields (CFG-FF / CFG-FN), participants (CFG-PT), and security roles (CFG-RS). Every deviation cites a specific rule identifier you can read and challenge.

A health score you can interrogate

The domain score out of 100 is computed from weighted indicator ratios with published green/amber/red thresholds. The full methodology table (observed value, band, threshold, weight) is printed beside the score, so a board or a regulator can check the arithmetic.

roadmapOption-set orphan detailCFG-OS

Option-set entries referenced by no active field, duplicate entries within one set, and sets no live form still uses: dead weight that makes every configuration change riskier than it should be.

roadmapForm bloat per work typeCFG-FF · CFG-WT

Field counts per form per work type, with each form's share of dead and sparse fields, so you can see which intake forms grew longest and how much of each one is dead weight for the fee earner filling it in.

roadmapOrphaned configurationCFG-WT

Work types, aspects, and rules present in the tenant but referenced by nothing live.

2How we examine it

Configuration is captured through the administration API. This tier consumes configuration only, so no matter content and no personal data leave the tenant. Fill rates are computed over the live matter population; every standards deviation cites a specific rule identifier; and the health score's indicator values, thresholds, and weights are disclosed in full on the report.

Fill rates and counts are observed measures, and every standards-deviation call cites a rule identifier you can read and challenge. The score's weights and thresholds are disclosed, hand-checkable multipliers, a deliberate starting policy tuned with you per engagement, not a fitted statistical model. Dead-field detection is only as complete as the field catalogue supplied: a field with zero rows anywhere can be counted only when the catalogue export names it.

3Example finding

Illustrative example from the synthetic demonstration corpus

This estate scored 38.9 out of 100 on configuration health. Across 287 attribute fields measured over 30,000 matters, 189 fields (66%) carried almost no data: 95 never populated at all and 94 filled in fewer than 5% of matters. 1 field stored its values two different ways at once, and 5 logical fields were split across 10 raw keys differing only by case or punctuation.

  • 95 of 287 fields never populated at all: pure configuration debt
  • 1 field storing values two ways at once (free-text literal and option-set code), quietly splitting every report built on it
  • 5 logical fields split across 10 raw keys differing only by case or punctuation (CFG-FN)
  • Duplicate-field-name ratio landed in the red band: 1.7% observed against an amber ceiling of 1.5%
38.9 / 100configuration health score, methodology disclosed below
287attribute fields in the estate
189fields dead or sparse: 66% of the estate
30,000matters every fill rate is measured over
Core fields19 fieldsModerately used79 fieldsSparse fields94 fieldsDead fields95 fields

How the score is built, in full. Each indicator is a ratio of the field estate (so a boutique practice and a national firm are scored on the same scale), banded against published green/amber thresholds, then weighted. Nothing is hidden behind the number.

IndicatorObservedBandGreen / amber belowWeight
Dead-field ratio33.1%Amber< 10.0% / < 40.0%
Mixed-representation ratio0.3%Amber< 0.0% / < 1.0%1.5×
Duplicate-field-name ratio1.7%Red< 0.0% / < 1.5%
Read the full chapter. The sample report runs this lens end to end over the synthetic demonstration corpus: every finding, its evidence trail, and the scored methodology, in Configuration & field usage.
Fill-rate buckets across 287 configured attribute fields, with sparse and dead highlighted. Hover any bar for the exact count.

4The benefit

What you walk away with

A shorter, higher-signal form estate; reports you can trust; a health score the board can track quarter on quarter; and a prioritised clean-up backlog scoped to specific rule identifiers, so remediation is defensible to your board and, increasingly, to a regulator.

← All audit domains See it in a full report