What is a purchase outcome actually evidence of?
From MCC-001, an essay: a hypothesis, not a measurement.
The short answer
Not always what it appears to be. The essay argues that an outcome can depend on the supplier, the service, the agent and its configuration, and the task together, so a record can be accurate about what happened and still misleading about who produced it.
As published in the issue
The difficult part is deciding what an outcome is evidence of. A useful record would identify the supplier, the product or service, the agent and its configuration, and the task and context; an outcome may depend on their interaction. A record can be accurate about an outcome and misleading about who produced it.
The sentence above is quoted from the essay as published and is not re-written here. The essay, its date and the address it was published at are below.
Why it matters
A record that credits or blames the wrong party teaches every buyer who reads it the wrong lesson. Attribution is the hard part of a useful record, not the collection.
What we measured
Nothing here was counted. This page quotes the essay's argument about attribution, which the essay leaves open for an agent whose model or tools change. What our record publishes about endpoints changing their price or payment address is linked below.
What the evidence shows
- Issue
- Machine Commerce Conditions: MCC-001 — September 2026
- Published
Where our own record speaks to this
These are measured answers, each counted over its own population on its own date. This page links them and quotes nothing from them.
What this leaves open
If an agent’s model or tools change, how quickly should confidence in its previous record decay?
Method and reproduction
The terms
The terms this question uses are defined in the glossary.
Corrections and disputes
Corrections to this issue are on the corrections register, and this answer reads from the record rather than from a copy.