Methodology
Two separate steps: an estimate of the loss before optimization, and measurement of the result after it. They are not mixed together.
Estimate before optimization - the loss calculator
The calculator on the main page works with a conservative model: a below-average website converts about 1 % of visitors, a well-built website about 3 %. The difference, multiplied by monthly traffic and the average order value, gives an estimate of what the old website leaves on the table each month.
It is a deliberately conservative illustration, not a promise. Actual conversion varies by industry, product price and seasonality - the model shows the order of magnitude of the problem, not an exact number for your specific website. The exact number comes only from an audit of your website.
Measurement after optimization - two different horizons
Technical numbers (performance, accessibility, SEO, readiness for AI assistants) are measured right after launch with Lighthouse - they are objective scores, not a subjective impression. Both published case studies show exactly these numbers, mobile and desktop.
Conversion and revenue are a different category - they are measured continuously and over a longer horizon, not after the first week of operation. Short-term swings (seasonality, a single campaign) say nothing about the real effect of a redesign.
Limits
For both projects so far (bazarfoto.cz, fotobazoš.cz), conversion and revenue measurement has not been running long enough to give a statistically reliable number - so the case studies currently show only technical Lighthouse scores, not monetary impact. Once real data is available, it will be added to the same case studies, not to a new page with different numbers.