Incrementality

How it is measured

The standard method is a holdout test. Split a comparable audience or set of regions in two. Run the activity for one group, withhold it from the other, and compare outcomes over the same period.

Plain-text formulas:

• Incremental conversions = conversions in test group minus conversions in holdout group (with both groups scaled to the same size)
• Incrementality rate = incremental conversions divided by total conversions in test group
• Incremental cost per acquisition = channel spend divided by incremental conversions

A concrete case

For those companies, it argues, "incrementality testing becomes a more valuable tool than broad channel experiments." It also advises teams to "Validate channel incrementality before committing to budget renewals." The point of the exercise, in its words, is "separating real lift from coincidence."

Common mistakes

• Treating attributed conversions as incremental ones. Attribution assigns credit to touches. It does not show what would have happened without them.
• Using a holdout group that differs from the test group in region, season or segment. The gap then reflects the audience, not the spend.
• Running tests too short or too small to rise above normal week-to-week noise.
• Testing once and reusing the result for years after the channel mix and competitors have changed.

When it matters to you

It matters most at renewal time and in board conversations about efficiency. The same guidance describes a shift at scale "from growth-rate maximization to efficiency, measurement quality, and alignment across functions."

An incremental cost per acquisition figure gives the chief financial officer and the board a number tied to cause, not credit. Set it beside your spend ratio from marketing budget as a percent of revenue. Present it through the CMO board deck template.