Article
The Truth About Klaviyo's Attribution Window - YOCTO Agency
yocto.agency
Quoted on this wiki
Every place a page here uses this source, in the order the words come in it.
Pure-last touch: The sale is 100% credited to the last interaction the user had before purchasing. Simple and straightforward, but clearly falls short when using channels like Native Ads, which are discovery channels, and don’t have high levels of purchase intent. Most brands start here, but once they go multi-channel, the model starts having problems. “initiatives like marketing automation would be grossly misrepresented (as the first touch would be the marketing channel that brought the user to the website).” Even distribution:This model gives equal credit to every touchpoint in the journey. Fair? No. The point of attribution is to understand what’s most valuable, so by giving equal credit to all steps, you might be overattributing value to steps that are less important in the customer journey. Just as clearly, not every marketing impression has the same impact on conversion. Consider clicking a simple banner ad vs. consuming a long-form sales video.
The Untrackable Interactions “Regardless of your attribution model, much of what will influence your customers can and will fall through the cracks of attribution models.” Byron Sharp’s seminal work “How Brands Grow” highlights this and argues that you win by becoming top-of-mind (mental availability), not just through direct sales. Top-of-mind means that your brand is the brand people think about when they have a need for the type of solution you’re offering.
Pure-last touch: The sale is 100% credited to the last interaction the user had before purchasing. Simple and straightforward, but clearly falls short when using channels like Native Ads, which are discovery channels, and don’t have high levels of purchase intent. Most brands start here, but once they go multi-channel, the model starts having problems. “First touch: The sale is 100% credited to the first interaction a consumer had before eventually purchasing.” Even distribution:This model gives equal credit to every touchpoint in the journey. Fair? No. The point of attribution is to understand what’s most valuable, so by giving equal credit to all steps, you might be overattributing value to steps that are less important in the customer journey. Just as clearly, not every marketing impression has the same impact on conversion. Consider clicking a simple banner ad vs. consuming a long-form sales video.
Position-based: This model is U-shaped. 40% of revenue is attributed to the first touchpoint and 40% to the last touchpoint. The 20% that’s left is divided equally among the touchpoints in between. More logical than even distribution, yet, sadly, imperfect, as it would assign the same value to a free, search-driven visit, and an expensive retargeting ad, in the middle of the user journey. “This model gives credit to all touchpoints, but gives more credit to the interactions that happened the closest to conversion.” Data-Driven (DDA): Google Analytics 4 model, which employs statistical models and machine learning algorithms to evaluate every possible combination of touchpoints to determine their contribution to conversion. Good luck understanding how it works without a statistical background, especially since it’s largely a “black box,” that offers less transparency into how exactly the attribution decisions are made.
Time decay:This model gives credit to all touchpoints, but gives more credit to the interactions that happened the closest to conversion. So, assuming there were 10 touchpoints, the initial one would get… Well, it’s complicated, because how do you assign the correct weight to each step? How do you know that the shopping ad is more important than the podcast’s endorsement in pushing a user to complete a purchase, just because it happened later in the journey? “Data-Driven (DDA): Google Analytics 4 model, which employs statistical models and machine learning algorithms to evaluate every possible combination of touchpoints to determine their contribution to conversion.” The Untrackable Interactions
The Myth of ‘Correct’ Attribution “There’s no ‘perfect’ model, each has its flaws, whether it’s multi-touch or channel-specific attribution (like Klaviyo’s, or Meta’s ROAS).” Why? People hop around. They see an ad, maybe they get an email, maybe a friend mentions a brand, or maybe they listen to a podcast - and importantly, people don’t do any of these in the same order, at the same time, or in the same way for every order. The customer journey is rarely linear and funnel-shaped (from impression, to click, to purchase).
Pure-last touch: The sale is 100% credited to the last interaction the user had before purchasing. Simple and straightforward, but clearly falls short when using channels like Native Ads, which are discovery channels, and don’t have high levels of purchase intent. Most brands start here, but once they go multi-channel, the model starts having problems. “Tracking purchase decisions with longer cycles that involve many touchpoints would also be very problematic.” Even distribution:This model gives equal credit to every touchpoint in the journey. Fair? No. The point of attribution is to understand what’s most valuable, so by giving equal credit to all steps, you might be overattributing value to steps that are less important in the customer journey. Just as clearly, not every marketing impression has the same impact on conversion. Consider clicking a simple banner ad vs. consuming a long-form sales video.
Things are not clear-cut here either.There are various marketing attribution models, and, again, none of them is perfect. The simpler the model the less precise it is, and vice versa. Here are the most common ones: “Most brands start here, but once they go multi-channel, the model starts having problems.” First touch: The sale is 100% credited to the first interaction a consumer had before eventually purchasing. Sounds sensible, but if you were to do so, initiatives like marketing automation would be grossly misrepresented (as the first touch would be the marketing channel that brought the user to the website). Tracking purchase decisions with longer cycles that involve many touchpoints would also be very problematic.