Definition

The ratio of a candidate behavior's sub-cohort conversion rate to the baseline cohort's conversion rate — how much more likely a user who performed the behavior is to reach the commercial outcome, relative to baseline.

How it's computed

Compute the baseline first: the share of the signup cohort that reached the commercial outcome event within the analysis window. Then compute the same rate for the subset of the cohort who performed the candidate behavior. The multiplier is sub-cohort rate divided by baseline rate. A baseline of 5% and a sub-cohort rate of 30% produces a multiplier of 6x.

The baseline is computed once and reused across candidate behaviors, with one exception: when a candidate is structurally restricted to a subset of the population (web-only event, paid-tier-only feature, admin-only action in B2B), the baseline must be filtered to match the candidate's restriction. Otherwise the multiplier conflates the behavior's effect with the underlying conversion gap between the subset and the broader cohort.

What it measures

The multiplier is a per-user signal: how predictive performing the behavior is of an individual user converting. A 6x multiplier says users who did the behavior were six times more likely to convert than baseline users. It says nothing about how many users performed the behavior — a 6x on a behavior touched by 1% of the cohort is structurally different from a 6x on a behavior touched by 10%.

Pairing with percent of converters

Strength per user is half the picture. Multiplier analysis pairs each multiplier with percent of converters, which measures what share of actual conversions came from users who performed the behavior. A high multiplier with low percent-of-result is a niche lever — strong per user, limited reach. A low multiplier with high percent-of-result is widespread but weakly distinguishing, since most non-converters do it too. Both reads are needed; multiplier alone risks chasing levers too narrow to scale, percent-of-result alone risks optimizing behaviors that don't actually distinguish converters from non-converters.

Example

A B2B analytics tool runs multiplier analysis on its signup cohort against first paid conversion. The baseline conversion rate is 8%. Three candidate behaviors come back with these multipliers:

• Connected a data source in first session: 4x (32% converted)
• Invited a teammate: 7x (56% converted)
• Created a dashboard: 1.5x (12% converted)

The teammate invitation is the strongest per-user signal, but the multiplier alone doesn't say whether to prioritize it. If teammate invitations only happen for 3% of signups, the leverage is limited unless something upstream expands that reach. If dashboard creation happens for 60% of signups, even its modest 1.5x covers most of the conversion volume. The multipliers tell which behaviors predict conversion most strongly; reading them next to percent-of-converters tells which also have the reach to support activation work.

See also

• Commercial outcome anchoring