Boardroom Definition
Incrementality is a measurement methodology used to determine the specific percentage of business outcomes (sales, leads, traffic) that occurred solely because of a marketing intervention. Unlike standard attribution, which assigns credit to any touchpoint in a user's journey, incrementality seeks to isolate "incremental lift" the conversions that would not have happened if the ad had never been shown. It answers the CFO's ultimate question: "If we turned this budget off, how many sales would we actually lose?"
Incrementality is calculated by comparing the performance of a Test Group (exposed to ads) against a Control Group (withheld from ads).
The Lift Formula:
$$Incremental Lift = \frac{(Conversion Rate_{Test} - Conversion Rate_{Control})}{Conversion Rate_{Control}}$$
Incremental ROAS (iROAS):
To calculate the financial efficiency of the lift, rather than total revenue:
$$iROAS = \frac{Incremental Revenue}{Media Cost}$$
If a campaign generates $100k in revenue, but the Control Group shows that $80k would have happened organically, the Incremental Revenue is only $20k. If the media cost was $10k, the standard ROAS is 10:1, but the iROAS is 2:1.
The Real Scoop
In 2026, the battle between Attribution and Incrementality is the central conflict in media strategy.
Attribution is about correlation ("The user saw the ad and bought"). Incrementality is about causation ("The user bought because they saw the ad").
The "Insider" reality is that Retargeting campaigns often fail incrementality tests miserably. Algorithms are designed to chase the "lowest hanging fruit" users who are already in the checkout flow. A standard report might show a Retargeting campaign with a 50:1 ROAS, but an incrementality test often reveals that 90% of those users were going to buy anyway. The true value of media is found in Prospecting, where you are creating new demand, rather than just harvesting existing intent.
Watch Outs
- The Cost of Silence: To run a valid incrementality test (like a Conversion Lift Study), you must intentionally suppress ads for a portion of your audience (the Holdout Group). This "cost of silence" can be unnerving for brands who fear losing sales during the test period.
- Ghost Bids vs. PSAs: Modern testing uses "Ghost Bids" (logging where an ad would have served) rather than serving PSA ads (e.g., "Donate to Charity") to the control group. Ensure your platform supports Ghost Bids to avoid wasting media spend on irrelevant control ads.
- Geo-Lift Complexity: When individual user tracking is impossible (due to privacy laws), brands switch to Geo-Lift testing (showing ads in Ohio but not Kentucky). This requires rigorous statistical matching to ensure the two regions are actually comparable before the test starts.
