Illustrative worked example · not customer data

Reach fell 38%. Did the audience lose interest?

A generic explanation might say yes. The WITNESS method begins by treating the 38% drop as an observation—not the cause.

Observation

Total reach ↓ 38%

That change is material. By itself, however, it does not distinguish audience fatigue from distribution change, source-mix shift, campaign effects, reporting anomalies, or other explanations.

Illustrative evidence set

SignalChangeInterpretive pressure
Total reach−38%Something materially reduced exposure.
Discovery exposure−44%Supports a distribution/discovery explanation.
Returning audience+19%Contradicts broad audience-abandonment theory.
Completion+23%Audience that did receive the content consumed more deeply.
Core engagementUpFurther weakens simple “content lost relevance” explanation.

Competing explanations

A — Audience fatigue

Predicts weaker audience-quality signals alongside lower reach. The illustrative completion and returning-audience evidence pushes against this explanation.

B — Distribution shift

Predicts weaker discovery while the audience that still receives the work can remain healthy or improve. The illustrative evidence fits this better.

C — Reporting anomaly

Remains possible until source quality, reporting windows, and Signal Integrity are checked.

D — Source-mix change

Could explain lower total exposure with stronger core behavior if a low-intent discovery source contracted.

Illustrative Finding

The evidence does not support a broad decline in audience interest.
The stronger explanation is a distribution/discovery change: fewer people were exposed, while the audience that did receive the content showed stronger returning behavior, completion, and core engagement.

Uncertainty

This example does not establish the hidden cause inside any platform algorithm. WITNESS can infer a distribution shift from observable evidence without pretending to observe proprietary platform internals.

What would change the Finding?

If later evidence showed completion, returning behavior, and core engagement also deteriorating across reconciled periods—or Signal Integrity showed the apparent improvements were unreliable—the audience-fatigue explanation would gain support. That condition can become a Watch.

That is the difference between an explanation and an investigation.

The numbers above are deliberately illustrative. The purpose is to demonstrate the WITNESS reasoning standard: observation → reconciliation → competing explanations → challenge → Finding → uncertainty → change conditions.