Skip to content

Source Data Guide Pattern

02.030.020.040 v20260619.001

The Source Data Guide Pattern is a prescribed data-source pattern for identifying, qualifying, structuring, and governing source data required for a Practice Area, Practice, Pattern, Offering, proof artifact, model, or value realization engagement.

It helps Alescent move from vague requests for data to structured source discovery, data readiness, evidence design, and model governance.

Pattern Logic

A Source Data Guide should identify:

  • source data classes;
  • source data applicability;
  • source model requirements;
  • cost model requirements;
  • value model requirements;
  • readiness requirements;
  • Parameters;
  • Providers;
  • Points;
  • data readiness signals;
  • common defects;
  • validation questions.

Parameters, Providers, and Points

Each data source should include:

  • Parameters. Variables, options, thresholds, filters, dimensions, assumptions, or scoping conditions that can be manipulated or configured.
  • Providers. Roles, systems, functions, partners, platforms, or data owners that produce, maintain, approve, interpret, or control the source data.
  • Points. Individual facts, fields, observations, measures, records, events, or data elements used by the model or evidence process.

Applicability Scale

Symbol Applicability Meaning
blank Not Required Data source is not normally needed for the model, scope, or stage.
O Optional Useful where available, but not normally required.
R Recommended Strongly useful and should be requested where practical.
Q Required Normally required for defensible model, evidence, or delivery work.

Pattern Use

Use this Pattern when:

  • a Practice Area depends on recurring data sources;
  • data availability or data readiness affects delivery;
  • models require source traceability;
  • value evidence depends on partner data;
  • source systems and providers are unclear;
  • data quality defects are likely to affect conclusions;
  • the engagement needs a repeatable discovery guide.

Guardrails

  • Do not treat data availability as data quality.
  • Do not treat source data as authoritative until source authority is confirmed.
  • Do not omit data Providers when interpretation or access depends on roles or systems.
  • Do not omit Parameters when model outcomes change materially based on filters or assumptions.
  • Do not collapse Points into broad data-source names.
  • Do not require data that is not proportionate to the engagement scope.
  • 02/030/120/910/010.md