Teika AI Study: How Much Does an Attribution Window Matter?

October 1, 2020 | Posted By Andrew Waber
Teika AI Study: How Much Does an Attribution Window Matter?

This Teika AI column is authored by Chris Allen, Teikametrics Data Scientist

A question we get a lot at Teikametrics is whether or not a brand should focus on a shorter or longer attribution window when it comes to judging the conversions driven by an Amazon Sponsored Products ad. This speaks to the conflicting desire of sellers to see the longest-term impact of an ad, versus quickly making a decision based on the available data.

The good news for sellers is that after examining a year’s-worth of Sponsored Products ad data, we can say brands on Amazon should have a reasonable amount of confidence that the vast majority of conversions will be attributed within seven days. For conversions in all three multi-day attribution windows (7-day, 14-day, and 30-day), 90% or more conversions will arrive within a week, and in only four days for the 7-day window.

Results Point to a Very Skinny Long Tail

For conversions with 1-day, 7-day, 14-day, and 30-day attribution windows, we have reasonable confidence that most conversions will be attributed well before the attribution window ends.

Across all three multi-day attribution windows, 90% or more conversions arrive within a week, and in the case of the 7-day window, in only four days.

For clarity, here are the key metrics broken out:

What Sellers and Brand Owners Should Take from this Analysis

The results demonstrate that while it takes time for advertising metrics to settle on their final value due to latency in Amazon’s system, brands on Amazon can examine their metrics earlier. A 7-day attribution window is going to be very good at giving you a largely complete picture of how many impressions, clicks, and conversions will be attributed to your ads.

Knowing this, you should feel confident making decisions across a variety of strategic areas once the 7-day threshold is hit. Whether that’s deciding to increase your investment across a certain keyword, move a keyword from an automatic to a manual campaign, or whether your ad is driving the requisite conversion volume you had expected. Ultimately, this should help you set better guideposts when judging the value of your advertising over time.

—–

Teikametrics’ Data Science Team Mission

The use of machine learning technology is largely restricted to companies with incredible resources at their disposal. Teikametrics’ Data Science team aims to democratize the power of AI-backed decision making for ecommerce businesses by providing brand owners and sellers around the globe with transparent and understandable data-driven insights, recommendations, or actions.