Factor

One dimension you're testing (like CTA color or headline copy), with two or more variants.

A factor is one dimension of a campaign (CTA color, headline copy, page layout) that you want to test. A factor by itself is just a name; it becomes testable once it has two or more variants, the concrete values it can take (a "CTA color" factor might have "Blue" and "Lime" as its variants).

A campaign with exactly one factor and two variants is a plain A/B test. Add a second factor and you're into multivariate testing territory: that is also the only design that can surface an interaction effect between two factors, something no number of sequential A/B tests can see.

A factor's position among its campaign's other factors is fixed once cohorts have been generated from it: reordering factors afterward would silently change which combination each existing cohort actually represents, which is exactly the kind of quiet corruption a locked campaign version is meant to prevent.