A week that beats the last one feels like proof. At seed volumes it rarely is: below roughly fifty meaningful attempts a week, a good week and a bad week can come from the same unchanged process. Founders who read the good week as a signal spend runway on scaling decisions their process cannot yet support.

This talk redefines traction as a loop whose output you can predict from its inputs, and separates it from the spikes that founder heroics, lucky introductions and competitors' outages produce. It explains why runway is better counted in controlled attempts than in cash, and why a loop with flat output and low error is worth more than one with rising output and high error.

The audience leaves with three questions to ask before each week starts, and a measure of traction that survives small numbers.

Format: keynote or talk, 30–45 minutes; workshop version available on request. Audience: founders, investors, executives and ecosystem builders.