When you run ads, it’s easy to count how many people visit your site. But the real question is: how many of those visits happened only because of your advertising? If you run an audience holdout experiment, that’s what Cost per Incremental Site Engagement (CpISE) measures. It estimates the true cost of bringing in new visitors who wouldn’t have come otherwise.
Here's an analogy.
Imagine you're running a museum. Some people visit because they already love museums or happen to walk by. Others only come because they saw the museum's posters around town. CpISE helps marketers understand the cost of getting those extra visitors who showed up thanks to your ads.
To calculate CpISE, start with the incremental site engagement volume. That's the number of visits caused by ads, after subtracting the natural baseline behavior observed in a holdout group. Read this article to learn about incremental site engagement volume.
Here's the formula:
Imagine you spend $10,000 on an ad campaign. The ticket sales report shows 5,000 museum visits during the campaign. Using the standard cost per visit formula, you'd come up with a $2 per visit estimate.
At first glance, that rate might look efficient. But let’s apply the museum analogy. Out of those 5,000 visits, 3,000 were natural visitors, which means those people who would have come anyway, even without seeing your ad campaign. That leaves 2,000 incremental visits driven by ads.
Using CpISE, we come up with a different cost:
So, you’re paying $5 for each net new visitor driven by ads. The $2 standard cost per visit is misleading because it includes museum visits you would have received without paying for ads.
Identifies wasteful channels
Some channels show a low cost per visit, but high a CpISE. That’s because they mostly reach people who were already going to visit the site. In museum terms: you’re paying to advertise to people who already planned to walk in.
Justifies top‑of‑funnel spend
Awareness tactics like video or Connected TV (CTV) often look expensive with high costs per visit. But if the ads drive a large volume of incremental visits, like posters inspiring new museum-goers, CpISE can actually be lower than retargeting. This shows awareness media may be more efficient at pulling new people into the funnel.
Optimizes the customer journey
Comparing CpISE across different paths reveals where messaging sparks the most curiosity.
In museum math: Path A ads are much better at pulling in fresh faces than Path B ads.
Cost per Incremental Site Engagement measures the cost of net new site visits that advertising causes. It divides total ad spend by incremental site engagement volume, which excludes natural visits from a holdout group. Marketers use CpISE to understand true acquisition efficiency instead of inflated visit counts.
CpISE equals total ad spend divided by incremental site engagement volume. Incremental volume reflects visits caused by ads after the platform subtracts baseline behavior from a holdout group. This calculation isolates paid impact and avoids misleading averages that include free traffic.
CpISE differs because standard cost per visit counts all visits, including those that occur without ads. CpISE removes natural behavior and focuses only on net new engagement. This distinction explains why campaigns with cheap visits can still waste budget.
CpISE rises for retargeting when ads reach users who already intend to visit. Incremental volume stays low because few visits are truly caused by ads. Marketers often see low click costs but high CpISE in these scenarios.
CpISE helps marketers compare efficiency across different paths and channels. Lower CpISE paths create more new interest per dollar. Higher CpISE paths signal weak messaging or audience overlap, which can cause issues like a journey not spending efficiently.