Define the result before the system can optimize it.
The modeled study begins with a customer-defined primary result. Ecommerce accounts use contribution-adjusted purchases. Local-market accounts use qualified booked opportunities weighted by their expected value. Reach, clicks, leads, and platform-reported return remain diagnostic measures—not substitutes for the business outcome.
Every account receives a written goal definition, attribution window, qualification rule, budget boundary, and list of excluded outcomes before the evaluation period begins.
- Primary measure: verified results divided by total advertising dollars
- Secondary measures: cost per result, test velocity, budget volatility, and result quality
- Guardrails: minimum margin, geographic eligibility, frequency, and customer-defined exclusions
Compare systems, not isolated campaigns.
The illustrative cohort contains 48 ecommerce and local-market accounts observed for twelve months: 24 using Vyra-directed decisioning and 24 managed by experienced human teams. Accounts are matched by business model, prior spend, seasonality, historical result rate, and market complexity.
Human teams retain their normal creative, media, and landing-page processes. The Vyra cohort uses the same approved budgets and offers, while Vyra coordinates audience, creative, copy, placement, bid, and Vyra Pages experiments as one decision system.
- No account changes its primary result during the measurement window
- Budget increases require the same customer-approved thresholds in both cohorts
- Invalid leads, cancellations, refunds, and unqualified bookings are removed from both sides
Normalize the result so different businesses remain comparable.
Each matched human-managed account is assigned a result-per-dollar index of 100. Its paired Vyra-directed account is scored against that baseline after adjusting for the approved goal definition and verified outcomes.
Across the modeled cohort, the Vyra-directed index reaches 140.7. That represents 40.7% more customer-defined result per advertising dollar—not 40.7% more clicks, impressions, or platform-reported conversions.
- Vyra-directed result-per-dollar index: 140.7
- Human-managed control index: 100.0
- Illustrative median testing velocity: 3.8× the control cohort
State the limits as clearly as the modeled outcome.
This is future-state demonstration data, not a completed customer study or a guarantee of performance. A real study would require independent validation, complete cohort disclosure, statistical confidence intervals, and publication of material account exclusions.
Performance will vary with the offer, economics, creative inputs, market, conversion path, measurement quality, and customer-imposed constraints. Vyra should only publish the claim as an operating result after the production system and study satisfy those standards.