METHODOLOGY

Measured recommendations need a visible method.

Marketing Genius turns data into proposals and actions, but a score is not a result by itself. These methods define what is measured, what is estimated and where human review remains necessary.

AI visibility scoring

Prompt and source coverage

Measure how a defined set of prompts and answer surfaces represent a brand, competitors and category. Results depend on prompt set, date, geography and provider.

Creative scoring

Quality before performance

Evaluate creative against declared brand, destination and format checks. A quality score is not a prediction of commercial performance.

Competitor analysis

Observable signals

Analyse connected intelligence sources and qualify run duration, coverage and regional availability. Duration indicates continued use, not profitability.

Conversion experiments

Hypothesis to measured change

Record the page, hypothesis, traffic context, test window and outcome. Avoid calling an uplift verified without a defined comparison and sufficient data.

Attribution

Revenue context

Connect campaign and CRM signals where available. Attribution is model-dependent and should be read alongside source data, not as an absolute truth.

Forecasts

Assumptions first

Forecasts are estimates based on stated inputs. They must show assumptions, range and uncertainty rather than promise a guaranteed outcome.

EVIDENCE LABELS

A number earns context before it earns trust.

Verified customer results require source data and permission. Demonstration dashboards are labelled illustrative. Product benchmarks disclose conditions. Estimated impact states its assumptions. If the evidence is not available, the claim is not presented as verified.

Visit the Trust Centre