By Chris Novak, COO, Eden Collective
Once an ad campaign is live, every decision a marketer makes counts—yet many of those decisions are built on muddied metrics. The semantic confusion surrounding certain terms creates a gap between what looks good on paper and what actually drives performance.
Impressions and Sessions: The Illusion of Engagement
Impressions are a key element of advertising. Unfortunately not all impressions are created equally. In digital, an impression might mean a banner ad, while in TV, it’s tied to the broader exposure of a demographic.
Digital impressions are created when at least 50% of the pixels in a given creative are on the screen for two seconds. In contrast when buying on linear TV there will always be 100% of the ad shown for the amount of time that has been purchased —but no guarantee that anybody will be watching it. When it comes to OOH, estimates can be generated based on traffic volume around an asset.
Meanwhile, sessions— or moments where a user engages with your site—are often inflated by meaningless interactions, like an accidental click or a brief glance. Marketers risk chasing session numbers that have no real value, misallocating budgets toward platforms where the numbers look impressive, but the actual business impact is minimal.
CPA: A Misleading Guide
Cost per acquisition (CPA) is often hailed as a universal benchmark for success, but in practice, it varies dramatically across mediums. Let’s break it down: Cost, Acquisition, and how those correspond to media investments.
While it seems simple on the surface to just divide Cost by action, each component of the equation can carry plenty of nuance. Cost might seem self explanatory, but there are many considerations that come into play. When calculating cost, are you looking at just the cost of media or are there also addeed fees? If you are paying targeting fees or platform fees, how are those included? When do agency fees get added? There is plenty of variability when it comes to what counts as a “cost.”
The same goes for what counts as an Acquisition. In some cases acquisition is defined as one particular action, such as placing an order, whereas in other cases it entails multiple actions.
Then comes the question of how these actions connect back to media investments, which introduces further variability. When attributing a media impression to an action, there are several elements: the ad occurrence on a device, the connection of the device to an individual, and the methodology of how that impression relates to a particular action.
Sometimes the connection is deterministic, when for example a device ID is connected to an impression and an individual based on persistent identity. But in many cases the relation between these elements is probabilistic, which involves some degree of uncertainty. All of this is important for ensuring that everything is apples to apples, and they often are not when switching between platforms.
On linear TV, CPA can be measured in many ways—whether by immediate conversions after an ad airs or by IP address matchbacks within a 7- or 14-day window. In digital, CPA is typically stretched over longer timeframes. They aren’t really the same.
Why is this important? CPA is a trademark of every platform. Some platforms might choose to obfuscate things in order to further their business. It’s important to go through the list of items that are included/excluded in the measurement so that you can make the right decisions.
The other thing about CPA is that it’s just a heuristic. Try talking to your CFO about the amazing CPA you are seeing on Facebook. While the costs are real, the actual results may or may not reflect the business outcomes they care about. Marketers that don’t acknowledge the fact might find themselves in an embarrassing situation where they have to explain why they continue to pump dollars into a channel that shows a strong CPA while overall sales are dropping.
Further, marketers often shift budgets toward digital from linear TV after seeing favorable CPA numbers, mistakenly believing that performance will be replicated across platforms. And when brands divest from linear TV entirely, expecting digital CPA to deliver equivalent results, they risk undermining long-term growth. In some cases, this misalignment can be so severe that it collapses entire strategies, sending brands on a path to short-term gains that are totally unsustainable.
Lookalike Audiences and Match Rates: The Fantasy
Cross-platform identity resolution and lookalike audiences are sold as a precise science. The reality? It’s far from foolproof. Match rates are often inflated to create the illusion that you’re accurately targeting users across devices, when in truth, the process is probabilistic at best.
Performance marketers, relying on this shaky foundation, often over-credit mobile and desktop environments because their identity resolution seems “perfect,” while under-crediting CTV or larger screens, where device matching is weaker. This overconfidence distorts optimization efforts, pulling spend away from where it can drive real results.
There is no “right answer” to many of the questions posed here. Rather than provide fixed definitions for these terms, the main point is to understand that they can mean different things depending on the campaign, the context, the platform, et cetera. The point is to make sure you ask questions so as to arrive at common definitions in the context of your campaign, your collaborations, and in defining your own success.
