Skip to content
November 9, 2023

The Truth About Klaviyo’s Attribution Window

Today, we’re telling the truth about Klaviyo’s attribution window. We’ll keep it simple and to the point. Klaviyo’s approach is transparent, flexible, and fair. Their technology gives you control over how you measure success in ways most other ESPs can’t (or prefer not to). 

Let’s dive right into it. 

Klaviyo’s Attribution Flexibility

By default, Klaviyo uses a 5-day attribution window that starts after a recipient opens or clicks an email. Obviously, since Apple’s Mail Privacy changes, you can’t always trust opens. But you don’t have to. Klaviyo lets you filter Apple Privacy opens out. Moreover, they allow you to shorten or expand the conversion window from 1 day all the way up to 30 (though, as you’ll see next, there’s a reason why it’s 5 days). 

You’re also not limited to measuring success by opens. You can easily switch to click tracking. 

It’s important to understand that these are not standard options with many other ESPs. Most ESPs won’t even tell you what their attribution window is unless you ask. There are email platforms that lock you into a set window, which is often large, like 30 days. For example, Active Campaign starts you on a 7-day window and lets you choose 14 days, 30 days, 90 days, or even 180-day windows. Other technologies, like Drip, completely disregard email engagement. If someone receives an email, visits a website, and then makes a purchase within 5 days, they will claim the sale. 

That’s like giving your weather app props for a sunny day just because you checked it in the morning. 

Moreover, if you’re using more than one Klaviyo marketing channel (e.g., email and SMS), Klaviyo uses cooperative multi-channel attribution. The sale is attributed to whichever message was clicked or opened most recently, as long as that conversion also falls within that message’s conversion window.

Suppose a user receives an email on day 1, then clicks an SMS on day 2, and purchases. Revenue will only be attributed to SMS if they purchase within 24 hours (which is the SMS channel’s default attribution). If they rather purchase on day 4, the sale will be attributed to email.

This is a powerful, easy, and fair way to evaluate how the two channels perform, and is miles ahead of the attribution you’d see in, say, an SMS-only platform. 

The Myth of ‘Correct’ Attribution

The idea that there’s one ‘true’ way to track where sales come from is a myth. 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). The latter type of attribution, by definition, ignores every other marketing touchpoint, so, clearly, it can’t give you a full picture. 

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). 

Google’s 98-page paper, Decoding Decisions, proves this conclusively and is worth a read by every marketer who takes themselves seriously. There’s a “messy middle” between interest in a product and buying a product, a loop of exploration (the expansion of options), and evaluation (the narrowing down of options)

In other words, shoppers are rarely only influenced by one single channel, and they rarely only buy from your brand (to think otherwise is downright silly). 

But What About ‘Multi-Touch’ Attribution Models?

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:

  • 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. 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. 
  • 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.
  • 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. 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.

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. It’s the chatter about your brand that doesn’t come with a hashtag or a brand name that comes to mind at the moment of need. None of these can be neatly filled into a spreadsheet, yet they nonetheless influence your brand’s growth.

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.

So even though you can’t track every interaction, what you can do is simply be there, ready to be seen at the moment of need. Relating this to CRM, every email that lands in someone’s inbox counts as a brand impression. They might not open or click on it right away, but that doesn’t mean it won’t play a part in their choices when they’re ready to buy.

So What’s A Brand Left to Do?

Let’s bring this back to tracking Klaviyo’s performance. The first thing you should do (especially when starting to work with an agency like YOCTO) is to define how you’ll measure success. There’s simply no sense in measuring performance in one way, and then suddenly changing it a couple of months down the road. You will no longer be able to make historical, apple-to-apple comparisons, so decide and stick to one way of measuring success. 

In-Platform Klaviyo Revenue Tracking:

 If you decide to use Klaviyo’s reporting to track revenue, we recommend switching to a clicks-only, 5-day conversion window. Yes – the revenue share % will go down substantially, but it’s illusory, so don’t panic. Your business is doing exactly as well as it did before, you simply switched to a new way of interpreting your performance.

  • Why clicks only? Because clicks are a much greater signal of interest and intent than opens and are ‘fake’ significantly less frequently. 
  • Why 5-days? Memory recall drops by 50% 24 hours after seeing an ad but remains at 50% for the next 5 days before dropping further, according to Nielsen research. A conversion window more narrow than that would not include all sales, and a conversion window larger than that would over-attribute sales, as it’s unlikely users would be influenced by something they did not remember.

Your mindset: Stop thinking in terms of ‘creating revenue’ and start thinking in terms of ‘contributing to generating revenue.’ That’s a slightly less attractive narrative than “we generate 40% of revenue via email” but it’s the truth.

Off-Platform Revenue Tracking:

If you decide to use Google Analytics 4, Shopify reports, or third-party platforms like Triple Whale to track email revenue, you’ll end up having the following recurring discussion: “Why do we see something different here, than we do on Klaviyo?” Your response should be to send these people this article. Next, if people doubt the optimizations you run within Klaviyo, we suggest this:

  • Concede to track revenue off-platform. If the entire organization decides to track email revenue in another platform outside of Klaviyo, then there’s a low chance they will let you do so, so it’s better to concede that. 
  • Educate your colleagues. Particularly in how no attribution model is correct, how each of them will misrepresent some interactions, and the concepts of winning on mental availability and untrackable interactions.
  • Focus on growing unique email clicks. Why unique clicks? So you get a clear picture of individual interest and intent without the “noise” of repeat interactions. An increase in unique clicks over time is a robust indicator of growing customer interest and intent. It suggests that not only is your content being seen by more unique individuals, but it’s also resonating enough to drive them to take action. 

Your mindset: Shift your mindset from just ‘direct sales’ to building ‘mental availability.’ More clicks mean more minds thinking of your brand. Every email is an impression, even if someone doesn’t buy directly.

The core takeaway is this:

There is no ‘true’ attribution model that captures every interaction, let alone capture it correctly. Klaviyo’s out-of-the-box tracking is fair, transparent, and very flexible. For most brands, our recommendation is to track clicks, and not opens, sticking to the 5-day window, as memory recall is relatively stable during the first 5 days of having been exposed to a marketing message.

Remember ‘direct sales’ don’t tell the whole story of your brand’s growth. How present you are in the minds of your audience the next time a need for your product materializes, plays a much greater role.

Consequently, this means that just ‘showing up, consistently’ in their inbox can and will lead to growth, regardless of what your tracking system says.

Interested in Working with YOCTO?

Let’s build something extraordinary together.

Do you need
to maximize your

work with us