Before and after is not automatically causal.
Seasonality, selection, concurrent changes and measurement differences can produce an apparent improvement even when the intervention had no effect.
DTF therefore distinguishes the observed outcome from a causal claim. Where the evidence permits, the outcome can be compared with an appropriate baseline or control using a declared, reproducible method.
A decision-specific measurement record.
What came before
Define the reference population, period and exclusions.
When the outcome is observed
Use decision-appropriate measurement windows.
How comparison works
Retain the deterministic calculation and its version.
What the evidence supports
Missing windows or weak controls reduce confidence rather than fabricate a result.
Leading outcomes and economic outcomes mature at different speeds.
An operational signal can appear within days or weeks while the full economic result takes months. DTF records each against its appropriate window instead of claiming final value early.