Home / Metrics

What changes. What stays yours. What we can prove.

Compare three ways to engage CLAVA: one priority surface, a connected system, or a governed implementation. Each is shaped to the institution. The evidence below comes from a private reference implementation and stops at recorded local and static proof.

For potential clients: use this page when you need the measured evidence behind the approach before deciding whether a first review is worthwhile.

Measured July 11, 2026 · review was due August 11, 2026. The receipt remains inspectable, but freshness is not claimed.

Engagement depth

Three engagement depths — not three templates.

The method can repeat; the result cannot. Accessibility is an acceptance criterion at every depth. Acceptance evidence and deliverables are scoped to the engagement.

  1. 01

    Essence Foundation

    Best for
    One priority surface that needs a lean, accessibility-minded brand foundation.
    What changes
    Core type, color, layout, and interaction rules become a tested brand layer.
    What stays institution-specific
    Institutional identity, content, workflows, and editorial judgment.
    Public evidence shown here
    Tier weight, the recorded AA-floor gate, and an integrity-bound source receipt.

  2. 02

    Signature Connected system

    Best for
    Related pages, programs, or instances that should feel like one school.
    What changes
    Shared patterns and relationships align the experience without erasing local ownership.
    What stays institution-specific
    Each team’s content, permissions, operating context, and voice.
    Public evidence shown here
    Tier weight, the recorded AA-floor gate, and an integrity-bound source receipt.

  3. 03

    Bespoke Governed implementation

    Best for
    High-complexity surfaces that need tailored patterns and a durable maintenance model.
    What changes
    Institution-shaped patterns, release evidence, rollback, documentation, and staff handoff.
    What stays institution-specific
    Institutional decisions, content, platform boundaries, and staff authority.
    Public evidence shown here
    Tier weight, the recorded AA-floor gate, and an integrity-bound source receipt.

Technical receipt

Measured brand-layer weight

Byte weight is one maintenance and performance signal; it is not the value of an engagement.

Essence
31.5 KB
Signature
32.7 KB
Bespoke
36.8 KB

Leaner is greener

We measure less, so teams carry less.

A directional transfer model, not a measured carbon inventory.

Leaner than prior

30.4%

Against a 52.9 KB prior gzip baseline.

Measured first-visit layer

39.7 KB

Four versioned text assets; HTML, binary fonts, and images excluded.

100,000 first visits

2 g

Estimated CO₂e at 4 GB transferred.

Accessibility evidence

Accessibility is an acceptance criterion — not a retrofit.

This snapshot records an enforced WCAG AA floor and complete computed-contrast accounting in the local and static test corpus. It does not certify a live client implementation.

Computed text and background

10,230candidates
100%accounted
0unknown exceptions

Across 80 fixtures and 223 viewport renders.

Content audit snapshot

23,194issues surfaced
86%deterministically auto-remediable
3,136require human judgment

Across 3,273 of 6,763 reviewed content records.

Proof boundary

Bound to local, static evidence.

The page says exactly where the evidence stops.

Proved here

  • Aggregate metrics match the imported artifact
  • Artifact bytes pass SHA-256 integrity verification
  • Tier and contrast gates passed for the recorded source
  • Private-data shapes are blocked from import

Not proved here

  • Live client or third-party runtime behavior
  • Owner approval or editorial sign-off
  • Publication, deployment, or uptime
  • Cryptographic authorship or keyed signature

Source receipt

Every figure has a source.

Integrity is machine-checked. Source authenticity is reviewed in the import PR.

Payload SHA-256
87e5109309454e61…
Source commit
ab714ce5
Schema
clava-public-metrics@1
As of
2026-07-11

Public snapshot · measured July 11, 2026 · review due August 11, 2026 · inspect the complete public artifact.

Have a similar problem? Request a fit conversation →