# Worked measurement exercise — fictional data

All URLs, periods and numbers below are synthetic teaching data. This is not an ai-fanout.com result, customer case or provider observation.

## Fixed scope

- Decision: whether a revised fictional SEO-tool buying guide needs further review.
- URL: https://example.com/seo-tools; English; Web; all countries; desktop only.
- Query cohort: fixed selection/cost reader tasks q01–q04 from the clustering exercise; recorded matching human query phrases would be required for real work.
- Baseline: one complete 28-day window before release.
- Follow-up: one complete 28-day window after confirmed recrawl; reporting latency allowed before export.
- Release: T0, adding a verified five-seat cost section.
- Recrawl: hypothetical T0 + 4 days. This is a planning assumption, not crawl evidence.
- Main metric: clicks for the fixed cohort; examine impressions and CTR alongside it.

## Synthetic arithmetic

| Window | Clicks | Impressions | CTR |
|---|---|---|---|
| Baseline | 10 | 1000 | 1% |
| Follow-up | 20 | 2000 | 1% |

Clicks doubled in this made-up dataset, but CTR did not improve. More impressions could reflect demand or broader visibility. No average-position value is provided: it cannot be derived from clicks and impressions.

## Change log and confounders

- T0: cost section and title changed together; effects cannot be separated by this comparison.
- T0 + 7: competitor changes and seasonal demand have not been checked.
- T0 + 4: hypothetical recrawl needs actual inspection evidence before treating it as fact.
- Native provider citations and query counts were not measured and are not inferred.
- Download use: unknown; do not insert a conversion rate without an existing permitted measurement source.

## Supported report

Observation: the synthetic follow-up has 10 more clicks and 1000 more impressions; CTR stays at 1%.

Hypothesis: the revised guide may cover more relevant questions. This remains untested.

Next action: confirm the query/filter scope and recrawl in a real review, inspect changed query mix and retain another comparable window. Do not report “fanout doubled traffic” from this exercise.