• George's Blog
  • Posts
  • Amazon's vCPM attribution model could be misleading your Ad reports

Amazon's vCPM attribution model could be misleading your Ad reports

Amazon Advertising Lessons - Oct 31

Last year, Amazon launched a new cost model distinct from the traditional cost-per-click (CPC): the cost per thousand viewable impressions (vCPM). This was introduced as part of the Sponsored Display optimization options - optimize for reach. Unlike CPC, with vCPM, you pay for every 1,000 views of your ad.

But there's more to vCPM than just the cost. It also has to do with how sales are attributed. This difference makes vCPM ads “appear” super profitable with big sales and low ACOS. However, this might not be the whole story. Some of those sales may not actually come directly from the ads but are still attributed to them. In simpler terms, the numbers might be bigger than they should be. This can lead to misinterpretations in your performance reports, potentially resulting in ill-informed advertising decisions and wasted Ad spend.

In this week’s email, let’s delve into this discrepancy, shedding light on how vCPM attribution functions and what needs to be considered while reporting them with other CPC campaigns.

Find out what’s wrong with your Amazon Ads. Book a PPC Audit.

The vCPM attribution and the conflict

In the case of traditional CPC campaigns, sales are solely attributed based on “clicks” on your ads. This implies that the sales generated by customers who clicked on your ad, visited the product's details page, and later made a purchase are attributed to the ad campaign.

With vCPM's view-based attribution, sales of your brand's products are attributed to both “clicks” and “views”. This means that if a shopper only views your ad and later makes a purchase of any product from your brand, those sales will also be attributed to the ad campaign.

So, even if a shopper was already planning to buy your product, whether they saw your ad or not, those sales will still be attributed to the vCPM campaign. It can potentially lead to overcounting of sales, making it seem like vCPM campaigns provide a high return on investment. In reality, these numbers should be viewed more as 'engagement' rather than true sales.

Figuring out the actual sales on vCPM campaigns

Taking into account the attribution of both clicks and views in vCPM campaigns, and the possibility of sales being overcounted due to the views attribution, determining the real sales, much like CPC campaigns, boils down to understanding the sales attributed to "clicks." These sales follow the same attribution method as other CPC campaigns and will help you plan the vCPM campaign optimizations and reports accordingly. There are 2 ways to find out clicks attributed to sales:

1. SD Advertised Product report

The SD advertised product report contains a breakdown of clicks attributed sales, in the form of additional columns including “14-Day Total Orders (#) – (Click)”, “14-Day Total Units (#) – (Click)”, “14-Day Total Sales – (Click)” “Total Advertising Cost of Sales (ACOS) – (Click)” as in below example:

As you download this report, you’ll find there are mismatches in sales and ACOS values in these columns, only in the case of vCPM, whereas an accurate match in the case of CPC campaigns (below example):

So, by downloading the SD advertised product report, you can get a breakdown of the clicks attributed sales and views attributed sales. Knowing this will help make more informed decisions in terms of spending on SD vCPM campaigns.

2. Targeting Tab

An alternative way, directly on the UI, is the targeting tab. This dashboard will show you only clicks attributed sales and the ACOS. You can use filters to select specific campaigns as well.

So, if you’re running vCPM campaigns, it’s worthwhile validating those numbers as the sales attribution model for vCPM ads, encompassing both clicks and views, raises doubts about the potential overcounting of sales, which could lead to a seemingly inflated sense of profitability.

Other learnings:

  1. The Products tab's Trends feature is a useful feature. It provides valuable insights into individual product performance, aiding in strategic ad planning. [Read more].

  2. Determining a product's % of total sales is crucial. It offers immediate insight into its impact, providing key clarity for strategic planning. [Read more].

  3. Amazon's AI for Sponsored Brands improves ad efficiency for large SKU lines. Useful for fast ad setup, where you can focus on top SKUs with professional creatives, and use AI for the rest. [Read more].

  4. New "Goals" settings for Sponsored Brands Campaigns. Now, you can select between driving page visits for brand consideration or increasing brand impression share for discovery and expansion. Note: Attribution is based on clicks and views. [Read more].

Thanks,

Himanshu

Read / Listen to last week's, or subscribe.

RESOURCES