--- title: "What Is Incrementality in Marketing?" description: "Incrementality measures the sales your advertising actually caused, not just what it got credit for. Learn how to measure it and use it to grow." author: "Samir Balwani" published: 2026-06-03 canonical: https://www.weareqry.com/blog/what-is-incrementality-in-marketing categories: ["Measurement & Analytics"] --- # What Is Incrementality in Marketing? The most expensive line item in most paid media budgets isn't a channel. It's the conversions you paid for that would have happened anyway. Every reporting dashboard is built to take credit. Platforms count the conversions they touched. Attribution models hand out points to the last click, or the first, or some weighted blend in between. None of them answer the question a CFO actually cares about: did this spend create new revenue, or just reshuffle revenue we were going to get for free? That question is what incrementality answers. And once you start measuring it, a lot of ["great" ROAS](/blog/why-a-high-roas-is-not-always-good) stops looking so great. ## What incrementality actually means Incrementality is the share of your results that happened only because of your advertising. The incremental conversions are the ones you would not have gotten if the ad had never run. Put simply: attributed revenue is what your ads got credit for. Incremental revenue is what your ads actually **caused**. The gap between the two is the money you're spending to reach people who were already going to buy. A retargeting campaign that re-converts shoppers already heading to checkout can post a 15x ROAS and be almost entirely non-incremental. A prospecting or [CTV campaign that introduces new customers](/case-studies/peak-design-ctv-geo-holdout-proves-incrementality) can look inefficient on platform ROAS while driving most of your real growth. ## Why ROAS and attribution miss it Platform-attributed and [last-click ROAS](/blog/ultimate-guide-to-marketing-attribution-for-paid-media) share the same blind spot: they can only measure demand they can see, and they instinctively claim demand that already existed. If someone searches your brand name and clicks an ad before buying, the platform books a conversion, even though that customer was already sold. Multiply that across branded search, retargeting, and loyal repeat buyers, and your blended ROAS quietly inflates while incremental growth flatlines. ## How to measure incrementality You can't read incrementality off a dashboard. You have to test for it by creating a group that doesn't see your ads and comparing it to one that does. There are five practical methods, roughly in order of rigor: 1. **Geo holdout tests**: Turn a channel off in a representative set of markets, keep it on elsewhere, and compare sales. Clean, board-ready, and hard to argue with. It's how we proved CTV drove real revenue for [Peak Design](/case-studies/peak-design-ctv-geo-holdout-proves-incrementality). 2. **Conversion lift / ghost-ad tests**: Platforms hold out a randomized control group that sees a placebo instead of your ad, then report the lift. Fast, but you're grading the platform's own homework. 3. **Audience holdouts**: Withhold ads from a random slice of an audience, such as a CRM segment, and measure the difference in conversion rate. Great for retention and retargeting questions. 4. **Time-based on/off tests**: Systematically pause and resume a channel and watch what happens to total sales. Blunt, but useful when geo splits aren't possible. 5. **Marketing mix modeling (MMM)**: A statistical model that estimates each channel's contribution from historical spend and sales. Best for top-of-funnel and offline media where user-level tracking breaks down. No single method is truth. The strongest read comes from [triangulating](/blog/triangulation-measuring-media-impact) two or three against each other. ## Incrementality vs. attribution vs. MMM These aren't competing religions; they answer different questions at different altitudes. - **Attribution**: Which touchpoints were involved? Good for day-to-day optimization, weak on causality. - **Incrementality testing**: Did this specific channel cause additional sales? The cleanest causal read, one channel at a time. - **Marketing mix modeling**: How does everything contribute together, including offline? Strategic, slower, directional. Used together they form a measurement stack: attribution for the daily steering wheel, incrementality for channel-level truth checks, and MMM for the strategic budget view. ## How to put incrementality to work You don't need a data science team to start. You need one clean test on the spend that matters most. 1. **Start with your biggest or most-doubted channel**: usually retargeting, branded search, or a new prospecting bet. 2. **Design a clean holdout**: a geo split or platform lift test with a control group that genuinely sees no ads. 3. **Run it long enough**: at least one to two full purchase cycles, so you're measuring behavior and not noise. 4. **Translate the result into incremental ROAS and CAC**, then compare it to your [break-even ROAS](/blog/calculating-a-break-even-roas-why-its-so-important), not your platform ROAS. 5. **Reallocate**: move budget toward what's provably incremental, and stop overpaying for demand you already own. Done consistently, this is what turns measurement from a reporting exercise into a growth engine: the bridge between efficiency and real, defensible growth. Incrementality is the discipline that keeps paid media honest. It's how you know whether you're buying growth or just buying credit for it. If you want help designing incrementality tests that hold up in front of your CFO, that's exactly what our [analytics and measurement](/services/analytics) team does. [Let's talk](/lets-talk).