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.
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.
Quality before performance
Evaluate creative against declared brand, destination and format checks. A quality score is not a prediction of commercial performance.
Observable signals
Analyse connected intelligence sources and qualify run duration, coverage and regional availability. Duration indicates continued use, not profitability.
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.
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.
Assumptions first
Forecasts are estimates based on stated inputs. They must show assumptions, range and uncertainty rather than promise a guaranteed outcome.
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