Skip to content
Product Metrics Assessmentpath:/services/

Independent judgement for decision safety.

Services are bounded assessment formats, each tied to a distinct decision risk.

This is an independent assessment of whether your product metrics and behavioural signals can safely support the decisions being made. Each service exists to answer a single question.

This is for teams who already have analytics, but no longer trust what the numbers mean under scrutiny.

The deliverable is an assessment artefact: judgement, assumptions, interpretive dependencies, and failure modes — written to be defensible under scrutiny.

Request an assessment Trust & Fit

Services

Services are defined by the kind of decision risk being assessed — not by tools, artefacts produced, or hours of work.

If you already know the shape of your problem, pick the closest service shape. If you do not, the canonical Product Metrics Assessment will be a good starting point.

Product Metrics Assessment

Default starting point.
Decision risk
Uncertainty about whether existing product metrics can safely support product, growth, or strategic decisions — especially when definitions are debated and confidence has eroded without a clear failure point.
Assesses
  • Metric definitions and stability over time
  • Alignment between stated intent and measured behaviour
  • Interpretive disagreement across teams
  • Dependence on undocumented assumptions
  • Exposure under internal or external scrutiny
Typically appropriate when
  • Metrics exist but feel unreliable or contradictory
  • Teams hesitate to make decisions based on dashboards
  • Definitions are debated rather than decisions being made
  • Confidence has eroded without a clear failure point

Signal Integrity Assessment

Decision risk
Loss of interpretability caused by upstream constraints that quietly alter what signals represent — including consent boundaries, session definitions, identity assumptions, and fragmentation across surfaces.
Assesses
  • Consent boundaries and the behaviour they hide or distort
  • Session definitions versus actual user intent
  • Identity assumptions across surfaces or time
  • Silent data loss or fragmentation
  • Aggregation assumptions that no longer hold
Typically appropriate when
  • Metrics appear technically correct but feel misleading
  • Numbers change without a clear product explanation
  • Teams suspect “we’re not seeing the full picture”
  • Decisions rely on behaviour that may not be fully observable

Decision Readiness Assessment

Decision risk
Metrics that function day-to-day but may not withstand increased scrutiny, accountability, or formal review — where confidence currently relies on informal knowledge.
Assesses
  • Defensibility of metrics under questioning
  • Clarity of assumptions at leadership or board level
  • Interpretive risk during audits, reviews, or transitions
  • Areas where confidence relies on informal knowledge
Typically appropriate when
  • Organisational scale or visibility is increasing
  • Leadership or ownership is changing
  • Decisions will be reviewed externally
  • Metrics must be justified, not just consumed

Metric Transition Assessment

Decision risk
Loss of meaning or false continuity during tooling, schema, or measurement changes — where “like-for-like” mappings create silent redefinition and historical interpretation risk.
Assesses
  • Continuity of metric meaning across systems
  • Implicit redefinition during migrations
  • Comparability and historical interpretation risk
  • Assumptions introduced by “like-for-like” mappings
Typically appropriate when
  • Migrating analytics platforms
  • Changing telemetry schemas
  • Consolidating reporting sources
  • Re-baselining metrics after change

Scope

Read Methods →

Each service is time- and scope-bounded. The engagement completes when the assessment artefact is delivered.

This page describes the service shapes; the engagement scope is defined during scoping and captured in the assessment’s scope statement.

Follow-up validation is a separate assessment with its own scope and artefact.

  • Deliverable is a written assessment artefact (judgement, assumptions, interpretive dependencies, failure modes).
  • Inputs are minimal and determined during scoping (definitions, dashboards as evidence, telemetry descriptions, decision context).
  • Outputs are descriptive rather than prescriptive, and are written to be defensible under scrutiny.
  • Ownership of execution does not transfer.
Out of scope (by design)
  • Configure tools, instrumentation, or platforms
  • Redesign metrics or select KPIs
  • Build or improve dashboards/reports
  • Execute migrations, reconciliation, or backfills
  • Provide templates, frameworks, playbooks, or training
  • Provide ongoing analytics support by default
Execution ownership and implementation accountability remain internal.

Stop conditions

Read Trust & Fit →

Stopping is treated as a quality safeguard, not a failure mode. An assessment may be paused or concluded if it can no longer be conducted responsibly.

  • Required access cannot be provided.
  • Scope expands beyond assessment or implies execution ownership.
  • Independence cannot be maintained.
  • Political or organisational constraints prevent honest examination.

When services are not appropriate

Read Trust & Fit →

An assessment cannot be responsibly performed in every context. The constraint is usually decision ownership, observability, or access.

  • No meaningful telemetry or metrics exist yet.
  • The organisation is seeking implementation, optimisation, or ongoing analytics support.
  • There is no identifiable decision owner for the judgement being sought.
  • The goal is validation rather than examination.

Next pages

Methods
How decision-safety is assessed, and where metrics commonly mislead.
Trust & Fit
Fit conditions, expectations, and boundary protection.
About the reviewer
Who performs the assessment, and how independence is maintained.
Request an assessment
How contact works, what information is required, and what happens next.