Scope and boundaryThe model forecasts a range from a private sample. It does not estimate total platform query volume or claim that every answer appearance creates a visit.
4forecast stages
100%percentages shown with counts
3low, base and high cases
What the evidence supports
- A lift from 1 of 10 to 2 of 10 is ten percentage points and one additional observed appearance; both should be shown.
- Mention lift and citation lift require separate assumptions.
- Referral and conversion ranges should never be presented as observed results before measurement.
Forecast sequence
Each row depends on the prior row and must preserve its denominator.
| Stage | Formula | Required disclosure |
|---|---|---|
| Observed baseline | brand appearances ÷ fixed answer observations | Question count, providers, market and date |
| Target lift | target appearances minus baseline appearances | Rationale for the target range |
| Attributable visits | incremental cited appearances × observed or assumed referral rate | Observed versus assumed label |
| Commercial value | qualified visits × conversion chain × value | Low, base and high rates |
Avoid false precision
01
Show counts
Place the numerator and denominator beside every visibility rate.
02
Separate observed and assumed
Use formatting and source fields that make the boundary obvious.
03
Reforecast after evidence
Replace planning assumptions with measured rates as the program runs.
Methodology
- Defined four sequential forecast stages in the downloadable operating kit.
- Required counts alongside percentages and low, base and high commercial scenarios.
- Prohibited conversion of citations into clicks without an observed or explicitly assumed referral rate.
Limitations
- Private prompt samples do not reveal total provider demand.
- Answer variability increases uncertainty in small samples.
- Modeled commercial value is not an observed result.
Questions
What sample size is enough?
There is no universal number. Use enough stable, decision-relevant questions to avoid one observation dominating the result, and always report the count.
Can the model forecast no-click influence?
Only as a labeled scenario. Without an observable referral or controlled study, no-click influence cannot be attributed precisely.
Sources and data
- RadiusRank AI SEO tool: Current public product scope, three-day diagnostic and $79 month-to-month subscription, verified August 14, 2026.
- Google Analytics traffic acquisition: Official source and session acquisition definitions used in the models.
Found a changed price, product limit or source? Email team@radiusrank.com with the official URL. We preserve the snapshot date instead of silently rewriting historical data.