Amazon Attribution Gives Advertisers a Fuller Picture of the Path-to-Purchase

June 4, 2020, Melissa Reilly


Amazon Attribution Gives Advertisers a Fuller Picture of the Path-to-Purchase

June 4, 2020, Melissa Reilly

Amazon Attribution Gives Advertisers a Fuller Picture of the Path-to-Purchase

June 4, 2020, Melissa Reilly

Merkle Blog Image
Merkle Blog Image

Amazon Attribution Gives Advertisers a Fuller Picture of the Path-to-Purchase

June 4, 2020, Melissa Reilly

Amazon Attribution Gives Advertisers a Fuller Picture of the Path-to-Purchase

June 4, 2020, Melissa Reilly

Merkle Blog Image

Amazon Attribution Gives Advertisers a Fuller Picture of the Path-to-Purchase

June 4, 2020, Melissa Reilly

Merkle Blog Image
Merkle Blog Image

Amazon Attribution Gives Advertisers a Fuller Picture of the Path-to-Purchase

June 4, 2020, Melissa Reilly

One of the longest-running conversation topics in digital marketing is attribution – specifically, how do all of the touchpoints in a purchase path contribute to a conversion and, more importantly, how do you allocate your marketing dollars in a way that maximizes return given that information? Historically, factoring Amazon into the attribution picture was not possible because of limitations for tracking transactions on Amazon.com. Now, Amazon has a beta to attempt to solve that.

Amazon Attribution allows advertisers to add tracking to various campaigns to understand how users behave when they engage with other channels just before visiting Amazon. It’s available to both sellers and vendors, though there are some limitations for sellers that we’ll cover later. Once Amazon Attribution is set up, advertisers get access to on-demand reporting and, in some cases, audience insights based on the data collected by tracking tags.

Reporting helps advertisers optimize programs earlier in the purchase path

Amazon Attribution tracking can be applied to search ads, display ads, video ads, email, and social. There is, however, some nuance to what data is available for each type of program. Display, video, and email get the full range of reporting, including audience insights, click-through conversions, and view-through conversions. Search and social are limited to click-through conversions. Additionally, social clicks must drive to an Amazon landing page for attribution tracking to work.

Reporting metrics for Amazon Attribution cover a range of actions from awareness to purchase. From the traffic side, advertisers can see impressions, clicks/CTR, and unique reach. Once the shopper is on Amazon, advertisers can see detail page views, add-to-carts, orders, and sales, as well as the rate metrics associated with each.

The primary use case for this data that’s accessible to all advertisers using Amazon Attribution is to more heavily fund ad campaigns that eventually lead to an Amazon sale. Oftentimes platform performance is viewed in isolation, with its contributions to sales on other platforms ignored. If a particular channel or campaign is driving a lot of sales on Amazon, advertisers should adjust that campaign’s performance goals or budgets to account for that.

Limitations lessen the positive impact of Amazon Attribution on advertising programs
There are limitations in several areas of Amazon Attribution that, until addressed, will hinder advertisers from fully understanding the various channels and brand interactions that lead to an Amazon order.

1. Incompatibility with Amazon advertising: Currently, Amazon Attribution does not include touchpoints from Amazon advertising and does not deduplicate against orders driven by ads. Let’s say a customer clicks on a paid search ad for “daily planners” that lands on advertiser ABC’s own domain, browses around for a while, and ultimately bounces from the page. A few days later, she goes to Amazon.com, searches for “daily planners”, clicks on advertiser ABC’s Sponsored Product ad, and then purchases on Amazon. That order would be double-counted – once as a paid search-driven order, and once as a Sponsored Product-driven order.

2. No multi-touch attribution: Currently, only the last marketing touch in the purchase path prior to a customer buying on Amazon is recorded. This could lead to undervaluing the most upper-funnel tactics that start the customer journey, while overvaluing the last click prior to purchase. This is especially troublesome for brands selling higher-ticket items with a longer consideration timeline leading up to purchase.

3. Fourteen-day attribution window for all advertisers: This isn’t likely a large hurdle for many brands. However, for categories with very short conversion windows, such as fresh flowers or gift baskets, or long conversion windows, like furniture, a strict fourteen-day window is not the ideal timeframe for viewing attribution data.

4. Limitations for sellers: From a reporting standpoint, sellers do not get access to audience insights and view-through conversions, while vendors do. Additionally, Attribution can only be tracked for sellers on ads that link to an Amazon landing page, like a store or product detail page.

The bottom line: A good first step, but we’re excited for the next iterations

Advertisers will certainly benefit from having insight into how other channels drive traffic to Amazon, and the tool in its current iteration is a big step in the right direction. However, more advanced advertisers and agencies will get the most impact from Amazon Attribution if the tracking capabilities are expanded and the tool’s constraints become more flexible. It’s critical that the entire purchase path gets illuminated with multi-touch attribution, including customers’ interactions with advertising on the Amazon platform. The other funnel behavior missing is what happens when Amazon is part of the purchase path, but not where the conversion ultimately takes place. While this is not intuitive purchasing behavior, that data would unlock another piece of the attribution puzzle to paint a more complete picture of how brand interactions influence conversions, whether they occur on Amazon or off.

You might also be interested in:


You might also be interested in:

You might also be interested in:

You might also be interested in:

You might also be interested in:

You might also be interested in:

You might also be interested in: