Intended use
What this kit is being designed to do
- Audit founder data
- Define business metrics
- Compare transparent forecasts
Business mini-lecture
Turn startup data into decisions, not dashboards.
Startup data becomes dangerous when a polished chart hides mixed definitions, missing records, or a changing denominator. A founder-data-analyst capability audits the dataset, defines each metric, and makes transformations reproducible. It becomes valuable once customer, product, or revenue activity repeats often enough that memory and ad hoc spreadsheets can no longer explain what is changing.
The agent can check data quality, calculate funnels and cohorts, compare scenarios, and show where a forecast depends on assumptions. The founder still chooses the business question, approves metric definitions, and judges whether a pattern is decision-worthy. Statistical consistency cannot decide whether to change positioning, invest in retention, or accept the risk of a sparse sample.
Operating contract
Inputs, outputs, and gates
Inputs
- Versioned dataset
- Decision question and metric definitions
- Analysis constraints and risk boundary
Outputs
- Data intake audit
- Reproducible analysis
- Decision memo with uncertainty
Gates
- Founder approves metric contracts
- Founder reviews transformations and limitations
- Founder makes the final decision
Known limits
What this record does not promise
- This is a roadmap placeholder, not an implemented analytical runtime.
- The scout did not establish fitness for production data or autonomous recommendations.
Lineage
Where the intended practice comes from
Derived from the validated 2026-07-29 decision-analytics scout bundle; implementation and runtime evaluation remain future work.
Lineage explains provenance; it does not substitute for a review record, field evidence, or a versioned implementation.