If you’ve spent any time in Google Ads, Meta Ads Manager, or a marketing analytics dashboard, you’ve seen the word “conversions” everywhere. It shows up in reports, in optimization targets, in the numbers your team presents in weekly meetings. But here’s the uncomfortable truth most marketers eventually run into: not all conversions are created equal, and the number sitting in your dashboard often isn’t telling you what you think it’s telling you.
This is where the distinction between incremental conversions and attributed conversions becomes critical. They sound similar. They’re often confused for one another. But they answer two completely different questions, and mixing them up can lead to some very expensive decisions.
This article breaks down exactly what each term means, how they’re calculated, why they frequently disagree with each other, and how to use both correctly instead of picking one and ignoring the other. If you want the broader context on how this fits into a full measurement strategy, our Complete Guide to Conversion Lift Measurement is a useful companion to this article.
Also Read: How to Get Your Google Ads Lift Study to 90% Certainty
What Are Attributed Conversions?
Attributed conversions are the conversions your ad platform or analytics tool credits to a specific marketing touchpoint, such as an ad click, an impression, or an email open, based on a set of attribution rules. When you log into Google Ads or Meta and see a conversions column, you’re looking at attributed conversions.
The key thing to understand is that attribution is a rules based system, not a measurement of causation. A platform looks at the sequence of touchpoints a user interacted with before converting, applies a model such as last click, first click, linear, or data driven attribution, and then assigns credit accordingly. The conversion gets counted whether or not the ad actually caused the person to buy.
Here’s a simple example. Imagine someone already intended to buy your product. They saw your Instagram ad, ignored it, searched your brand name a week later, clicked a paid search ad, and completed the purchase. In a last click attribution model, that entire conversion gets credited to the search ad, even though the person may have converted anyway without ever seeing it.
That’s not a flaw exactly. Attribution was never designed to prove causation. It was designed to help marketers understand which channels and touchpoints appeared in the path to conversion, which is genuinely useful for budget allocation and channel level reporting. The problem arises when attributed conversions get treated as proof that advertising worked, because that’s a claim attribution data simply can’t support on its own.
What Are Incremental Conversions?
Incremental conversions represent something fundamentally different. Instead of asking which touchpoint should get credit, incrementality asks a much harder and much more valuable question: how many of these conversions would not have happened without the marketing activity?
This is a causal question, and answering it requires a different kind of methodology than attribution modeling. Incrementality is typically measured through controlled experiments, such as conversion lift studies, geo based holdout tests, or user level randomized controlled trials, where one group is exposed to advertising and a comparable control group is not. The difference in conversion rates between the two groups, adjusted for statistical noise, is your incremental lift.
Going back to the earlier example, if that same brand search shopper would have converted regardless of seeing any ads, a well designed incrementality test would reveal that the campaign contributed zero incremental conversions for that particular customer, even though attribution handed the credit to a search ad. Incremental conversions strip away the noise of people who were going to convert anyway and isolate the conversions that are genuinely additional, meaning they only happened because the marketing existed.
This is why incrementality testing is considered the gold standard for proving whether a campaign is actually working, rather than simply showing up in the path to a conversion that was likely to happen either way.
Why Attributed and Incremental Numbers Rarely Match
If you’ve ever run a lift study and compared the results to your platform’s reported conversions, you’ve probably noticed the numbers don’t line up, and sometimes the gap is enormous. This isn’t a bug or a sign that something is broken. It’s the expected outcome of two systems measuring fundamentally different things.
Attributed conversions tend to overstate true impact. Because attribution models assign credit based on touchpoint presence rather than causal contribution, they frequently overcount the effect of channels that reach people who were already likely to convert, particularly branded search, retargeting, and other lower funnel tactics that tend to catch warm audiences right before they take action anyway.
Incremental conversions are almost always lower, sometimes dramatically so. It’s common for a lift study to reveal that only a fraction of attributed conversions are truly incremental. A campaign showing 1,000 attributed conversions might reveal only 300 to 500 truly incremental conversions once a proper holdout comparison is run. This doesn’t mean the campaign failed. It means attribution was crediting conversions that would have happened anyway.
Cross device and cross platform tracking gaps distort attribution further. A person might see an ad on their phone and convert later on a laptop, or convert through a channel your tracking pixel never captured. Attribution systems fill these gaps with modeled estimates, which introduces additional distance between attributed numbers and what actually happened.
Privacy changes have widened the gap even more. With the decline of third party cookies, restrictions on mobile device identifiers, and platforms relying more heavily on modeled conversions, attributed numbers have become increasingly approximate rather than precise, making the case for incrementality testing even stronger.
A Practical Example That Shows the Difference
Imagine a mid sized ecommerce brand running a retargeting campaign aimed at people who visited the site but didn’t purchase. The platform reports 2,000 attributed conversions for the month, and on the surface that looks like a strong result worth scaling further.
The brand then runs a holdout based incrementality test, excluding a portion of the retargeting audience from seeing ads entirely. When the results come back, the exposed group converts at a rate only slightly higher than the holdout group. After running the statistics, the true incremental lift comes out to roughly 400 conversions.
What happened to the other 1,600 attributed conversions? Most of those people were already planning to return and complete their purchase, retargeting or not. The ad simply happened to be present in their path, and the attribution model gave it credit anyway. This doesn’t mean retargeting is worthless. A lift of 400 incremental conversions might still justify the spend. But it completely changes the return on investment calculation, and it prevents the brand from over allocating budget to a channel that looked far more powerful on paper than it actually was in reality.
When to Use Attributed Conversions
Attributed conversions still serve a real purpose, and dismissing them entirely would be a mistake. They’re useful for day to day optimization, since most ad platforms use attributed conversion data to power automated bidding and budget allocation in real time, and incrementality testing simply isn’t practical to run continuously at that granularity. They’re also useful for understanding the customer journey, since attribution data shows which touchpoints tend to appear together in a conversion path, which is valuable for planning creative sequencing and channel mix. And they’re useful for fast, directional feedback, since attributed conversions update quickly and can flag obvious problems, like a campaign that suddenly stops generating any conversions at all.
When to Use Incremental Conversions
Incremental conversions become essential in a different set of situations. Budget allocation decisions at the strategic level should lean heavily on incrementality data, since deciding whether to invest more or less in a channel is fundamentally a causal question that attribution alone can’t answer reliably. Proving marketing effectiveness to leadership also calls for incrementality, since executives and finance teams increasingly want proof that marketing spend is causing business growth rather than simply riding alongside it. And auditing channels prone to attribution inflation, particularly branded search, retargeting, and affiliate marketing, is another clear use case, since these channels are the most likely to show attributed numbers that dramatically overstate true causal impact.
How to Bring Both Together
The most effective measurement strategies don’t treat this as a choice between one metric or the other. They use attributed conversions for speed and operational decision making, while periodically validating those numbers against incrementality tests to keep the overall picture honest.
A practical approach looks something like this. Continue using attributed conversions for daily and weekly optimization, since platforms need that signal to run automated bidding effectively. Run incrementality or lift studies on a recurring basis, such as quarterly or whenever a major budget decision is on the table, particularly for channels suspected of attribution inflation. Use the results of incrementality tests to build channel level adjustment factors, essentially discount rates that translate attributed numbers into a more realistic estimate of true incremental value. And revisit those adjustment factors periodically, since the gap between attributed and incremental conversions can shift as tracking technology, privacy regulations, and consumer behavior continue to evolve.
The Bottom Line
Attributed conversions and incremental conversions aren’t competing metrics where one is right and the other is wrong. They answer different questions. Attribution tells you where a conversion showed up in someone’s journey. Incrementality tells you whether your marketing actually caused that conversion to happen in the first place. Relying exclusively on attributed numbers risks overinvesting in channels that look effective but are mostly capturing demand that already existed. Relying exclusively on incrementality testing, on the other hand, sacrifices the speed and granularity needed for daily optimization. The smartest marketing teams use both, understanding exactly what each one can and can’t tell them, and building a measurement strategy that reflects that honest distinction.
For a deeper look at how incrementality testing is designed and executed in practice, revisit our Complete Guide to Conversion Lift Measurement, which walks through the methodology behind the numbers discussed in this article.