Almost all valuable data resulting from ad campaigns hinges on good attribution. If you don’t know which ads are effective, any other data pulled is of limited use.
Amazon defines attribution as the assigning of success to an ad that a customer was exposed to before taking a desired action, such as a purchase. This article will explore the ad types available on Amazon, how attribution can vary depending on ad type, and highlight the different reporting functions. It will also explain how to use Amazon attribution beta, and why it needs to be looked at in a wider context and need for analytics tools to help deliver this analysis and insight.
PPC and other paid advertising
Ad types provided by Amazon Advertising break down as: those available to Sellers and Vendors exclusive to Amazon; those available to larger retailers who would sell on both Amazon and on other sites; and those who don’t even sell on Amazon (e.g. telecoms providers or financial services companies) but would like access to Amazon’s user base.
For Amazon Seller and Vendor advertisers, there are three main PPC (pay per click) ad types available — Sponsored Products, Sponsored Brands and Sponsored Display.
Sponsored Products and Sponsored Brands are keyword related — Sponsored Display is more targeted at buyer behavior, and is a banner ad essentially. You choose your PPC ad type dependent on what your goals are: Product — targeted product sales; Brand — general product sales and awareness; Display — awareness and general product sales, along with retargeting.
Ad types availability is also dependent on your Amazon retailer level. Sponsored Products are available for Sellers, Vendors and agencies. Products must be in a category that is ads eligible and be eligible for the Buy Box in order to advertise. Sponsored Brands and Sponsored Display ads are available for Sellers enrolled in the Amazon Brand Registry, Vendors and agencies. There are also some regional variations — so make sure to look into your market specifics.
How they work
For Sponsored Products, you choose keywords to target or let Amazon’s systems target keywords automatically. You control how much you want to spend on your bids and budgets. Sponsored Brands (which can feature up to three relevant products) are also keyword-targeted ads. They are displayed at the top of search results (best position – highest cost), on the left-hand rail, and the bottom of page of search results.
Sponsored Brands can meet a number of goals, from generating awareness of a new product to promoting seasonal items or creating more demand for something selling well. Your choice of keyword approach tailors your ad to those goals — for example, generic unbranded short-tail keywords to attract new customers to branded keywords to defend your market share.
Sponsored Display is the most recent addition to the Sponsored Ads trio and adds “top of funnel” capabilities to target shoppers based on product categories, and “middle of funnel” capabilities to retarget audiences who showed interest in categories related to your promoted product but didn’t buy.
You can set up these re-engagement campaigns to reach customers browsing on or off Amazon. You can also create competitive campaigns by targeting the detail pages of specific products or product categories on Amazon.
The functionality of Sponsored Display is similar to the more advertiser savvy offerings in Amazon DSP, a separately priced demand-side platform. This provides automated, centralized media buying from multiple sources. DSP is part of the wider Amazon Advertising portfolio targeted at large spend advertisers whether they sell on Amazon or not. DSP, however, is different. For example, you would need to upload your own creatives, and it provides granular targeting options regarding geo-demographics and other behaviors.
Within the DSP console is a new function — Amazon Attribution. Still in Beta, Amazon Attribution is specifically aimed at tracking the performance of Amazon ads on channels off Amazon. The attribution measures page views, purchases, and sales driven by external campaigns. However, this is more strictly a traffic attribution tool. An email marketing campaign, for example, could generate traffic that would be captured here — rather than simply ad traffic.
For those with higher marketing budgets, Amazon also provides more customizable offerings including the ability to build video ads and customized display ads.
How Amazon attributes ad success
If you are a retailer on Amazon, you will at least want to compare how your Sponsored Brands ads perform against your Sponsored Product ads and potentially against your Sponsored Display ads.
If your goal is to drive brand awareness, you will want to pay close attention to your impressions; this will give you a sense of how much awareness you are driving with your campaigns. If your goal is to drive sales, you will look at your conversion rate to see how much of the awareness generated from impressions is converting into sales. You will also need to pay close attention to your average CPC (Cost-Per-Click), and here it’s important to understand that your CPC on Sponsored Brands can also vary for different placements: ‘Top of Search’ against ‘Other Placements.’
When a user clicks on one of your ads and buys a product within a certain time period, the “sale” is attributed to that campaign. Amazon uses a last-touch attribution model to try to take into account factors such as how the customer interacted with the ad.
In order to provide more specific attribution, Amazon has recently launched new-to-brand metrics, Amazon Advertising now provides display, video, and Sponsored Brands advertisers the ability to see whether an ad-attributed purchase was made by an existing customer or one buying your product on Amazon for the first time (over the prior year).
With ‘new-to-brand’, you can access campaign performance metrics such as total new-to-brand purchases and sales, new-to-brand purchase rate, and cost per new-to-brand customer. This tool will help estimate the cost of acquiring new customers on Amazon and identify the most efficient channels and tactics to achieve any ”new customer” campaign goals.
Another interesting aspect of attribution in Amazon advertising campaigns is “halo sales”. These form part of the sales attributed to a sponsored ad campaign, and are a good indicator that buyers are going for items other than the one that they originally clicked on. These sales also are referenced to the last ad a user clicked on.
If halo sales are included in your figures, this can have a positive effect on your ACoS — as Amazon will attribute more sales to your advertising campaign.
Halo sales attempt to measure the successful sales your ads have influenced across your whole brand portfolio and not just those you’re promoting. While halo sales are an important indicator of whether your ads are making an impact in other areas, you have to take into account how they are attributed and be aware of that in your analysis.
Limits of attribution
Unfortunately, Amazon tracks attribution differently depending on the ad types and platform you’re using.
For Vendors and Sellers, the Sponsored Brands attribution window is 14 days, meaning all transactions made within that time frame, after clicking on an ad, are added to your sales report. For those with API access, however, you can choose 7, 14 or 30 days for both Vendors and Sellers. For Sponsored Products, it’s 7 days and the advertiser must be the seller of the product in question. If the item is distributed by another retailer, Amazon does not include it in your campaign.
Sponsored Brands and Sponsored Display are similar, but the advertiser does not necessarily have to be the seller of the item. For halo sales, attribution windows are the same.
DSP also takes halo sales into account. However, they are not assigned based on ad clicks, but on visual contacts. The attributable period is 14 days.
Getting at the data
There are three different dashboards you can use to advertise on Amazon and therefore three different ways you can gather data on your ad’s performance.
- Seller Central is the dashboard third party Sellers use.
- Advertising Console, formerly Amazon Marketing Services (AMS), is the dashboard Vendors use (1P/first party sellers).
- Amazon DSP, formerly the Amazon Advertising Platform (AAP), run by the Amazon Media Group (AMG) is for larger advertisers.
The key takeaway is to be careful with your analysis: compare apples-with-apples, as much as possible, over the correct time frame. If you don’t, you can end up comparing one KPI, say advertising cost of sale (AcoS) for campaign types intended for completely different goals over too short a timeframe.
For example, brand awareness and conversions are two separate goals with different KPIs, and campaign measurements should take this into account. Amazon even recommends that advertisers have two separate teams or workstreams — one that focuses on top of the funnel, and a separate one that deals with bottom of the funnel.
What attribution on Amazon highlights is that before you build or alter your pay-per-click ad strategy, you will need to understand the data and how it is generated, especially when applying that to comparisons of cost per click, product conversion rates, or product profitability.
Attribution in a wider context
Knowing what has been bought is great, knowing who your customers are and what they are going to buy next is even better. Using Amazon data, you can build a far more detailed picture of who’s buying your products, where they’re from and which products they buy. When cross-referenced with persona categories, you go beyond simple attribution of sales to start to be able to see ‘buying trajectories’, and even calculate customer lifetime value.
Buying trajectories show the likely order in which products are purchased based on the persona and initial product of purchase. The challenge is wading through the data provided by Amazon to create a live picture. Trying to manage this with spreadsheets will ultimately fail. The future of such data-driven decisions can only be provided by AI-based analytics tools.
Customer lifetime value
If you can use data to understand the lifetime value of a customer, based on the products they are purchasing, the impact is huge. For example, you can not only look at that product’s specific margin, but at the value of acquiring that new customer and the ads that contributed to building that lifetime value. In a competitive marketplace, this can make all the difference. Again, this is a challenge that is only realistically surmountable using software and analytics tools. If you want to learn more, we recommend checking out our eBook on how to make sense of your Amazon customer data.
Digging into the data
Attribution and data analysis is about finding patterns. By understanding trends, you repeat success and mitigate failure. With enough data, you can even make accurate predictions about the future and take action to stay ahead of the curve. The less time your team needs to spend crunching the numbers, the more time is spent using data-driven insights to strategize and to create results.
The complexity of interpreting attribution data provided by Amazon is a prime example of why it is difficult to allocate enough time to uncover the many insights and opportunities that are buried in the data. If you want to stay ahead, it is becoming increasingly vital to have top-quality analytics tools to help transform your Amazon data into strategy and results. However, what is important is your ability to attribute actions so that you can double down on success.
Amazon will help you with attribution, but their internal tools only go so far. With that said, a great place to start is simply experimenting with a few campaigns. By actually putting products live and seeing what attribution results you get back, you will gain a better understanding of the kind of information available. Thinking about attribution from the beginning already puts you ahead of the curve. Make sure to keep up the good work.