Of the many advertising platforms around, Amazon has some of the most extensive and granular first-party consumer data. Having direct access to this type of data alone would make it easy for any advertiser to better target customers and improve their advertising performance. Sharing this information freely would violate customer privacy, which rightly so is a critical consideration for Amazon.
To empower advertisers without risking customer privacy, Amazon quietly launched the Amazon Marketing Cloud in late 2019. Since sharing identifiable customer data is a big privacy concern, Amazon took extra measures before launching the platform. Advertisers are able to access a wide array of retail and advertising data at the user level. However, they won’t be able to see or export sensitive customer data.
Now that we’ve set the stage for the privacy aspects, what is AMC really about? What are its capabilities? How come many advertisers haven’t heard of it and even fewer are actually using it?
What is Amazon Marketing Cloud (AMC)?
Amazon Marketing Cloud is a secure, cloud-based environment that advertisers use to perform and compare analytics across several pseudonymized data sets that enables them to create aggregated reports. Advertisers can contribute their own data (from CRMs, search data, or social media channels) when performing analytics with AMC. New data sources from Amazon are being added all the time – for example, Sponsored Brands is not even included yet – and there is a paid feature coming in Q1 next year that allows you to compare your advertising dataset vs benchmark datasets of shoppers who you have not reached with ads but who have interacted with your brand.
What does that mean in English? Its main purpose is to answer the question of how shoppers who interact with your campaigns are impacted by those advertising efforts and what subsequent actions they take such as purchases. It can help you start to understand concepts like Multi-Touch ROAS and various “cost per” metrics like New to Brand or Per Order.
Let’s break it down with a hypothetical: Joe is a frequent amazon.com shopper and is browsing news websites. He sees three Amazon DSP display ads while reading about current events. Two days later, he sees an Amazon DSP video ad for the same brand appear before a video he’s about to watch. Three days after that, he’s using keywords to search for products on Amazon. As he’s adding things to his cart, he sees several Sponsored Products ads from that same brand he’s now familiar with. He then clicks on one and buys.
Which ad gets the “credit” for this purchase?
Historically, Amazon has been a last-click attribution platform meaning the SP campaign would hog all the glory and leave those other campaigns looking useless, but with AMC you can start to build more complex models and distribute partial credit to various ads in the funnel, each of them a touchpoint.
Gone are the days when Amazon kept its datasets to itself. In fact, this level of data sharing might be even better than any other platform out there. With AMC, you can use these aggregate datasets as a playground to narrow down the data you really want to pull out key insights.
What the heck is a “privacy-safe clean room”?
Pseudonymization is a technique that replaces or removes information in a data set that identifies an individual but still allows you to understand detailed data at an individual level. AMC Is built with this privacy in mind such that the data is useful for advertisers but cannot be used to identify any one shopper as no PII (personally identifiable information) is accessible. Amazon frames this as a “clean room” like those used in the production of high-end electronics. Think of the people in full-body suits working in labs where there is no dust. It’s a bit like that, except for privacy. Despite a large and growing dataset being available in AMC it doesn’t mean that anyone using it can find out WHO they are observing the behavior of.
READ MORE: How does AMC contribute to Top-of-Funnel Advertising (TOFU)?
What are some new metrics you can find when harnessing AMC data?
Aside from the familiar ad-attributed clicks, impressions, and conversion metrics, Amazon Marketing Cloud has its own set of metrics that you can start to benchmark based on this new depth of data. These are just a few of the most important metrics. Each of them can be examined in a variety of ways such as at the campaign, ad type, or even target level.
NTE (New to Engage) – tracks everyone regardless of whether they’ve made a purchase or not as well as those who have engaged with your brand for the first time within the range of data you have available.
NTB (New to Brand) – tracks those who have purchased from your brand for the first time in the last year. You might be familiar with this metric from Sponsored Brand campaigns which have had it for a while.
MT-ROAS (Multitouch ROAS) – a unified RoAS considering all of the advertising types across Amazon DSP and Sponsored Ads. It takes into account the user journey with multiple touchpoints, each contributing a different amount to the final conversion.
LTV (Lifetime Value) – What value a customer has to your business over the whole period of their relationship with your brand. This looks very different for durable vs consumable products. Knowing this will help you determine what increases the value for existing customers vs what helps you acquire new ones and what the difference to the businesses bottom line is as a result. Also sometimes call CLV (customer lifetime value).
Why do so few advertisers use AMC?
“AMC sounds like an advertiser’s dream come true, but why are fewer people using it?”
AMC is under-adopted because people view it as a complicated analytical tool. This isn’t an entirely incorrect assumption but there are more and more companies that make working with this data source possible even if you don’t meet the technical skill and knowledge requirements yourself or have them in-house. If you have someone on the team who knows SQL or is capable of learning it you can start to extract basic insights.
Another reason is the lack of marketing muscle behind it. Amazon has not put their full weight into marketing this tool by talking up the incredible features it has and insights it can unlock. We mostly hear about it from other Agencies and people deeply involved in Amazon advertising on a day-to-day basis like ourselves. After the unBoxed 2022 event, this may start to change since much of the show was focused on mid and top-funnel advertising and quantifying those results.
A third reason is that getting access to an AMC instance is itself a bit difficult and for now requires at least some Amazon DSP ad spend. This will change in 2023 but for now, is the case and presents another barrier to entry for an advertiser using Sponsored Ads in the ad console alone.
The final reason is that AMC is still rapidly evolving and despite being around for a few years has some way to go before being truly complete. It doesn’t even have complete data from the different campaign types that we use to advertise on the Amazon marketplace.
So while these obstacles have slowed the adoption of AMC, it is becoming an increasingly valuable tool in an Amazon marketplace where complex buyer journeys, rising CPCs, and ever-expanding customer targeting options are in play.
What can you do with Amazon Marketing Cloud?
Amazon Marketing Cloud fills in the attribution gaps that advertisers struggle to understand when they rely solely on data from a single ad type. . Questions that advertisers struggle to resolve may be answered by AMC as it acts as a bridge between Amazon Ads and first-party signals. AMC offers several advantages centered mostly on the details of audience insights on the marketing, and advertising side. Here are some of those advantages:
Access event-level ad data
Amazon Marketing Cloud delivers something that all Amazon marketers and business owners have dreamt of for a long time which is shopper grain event-level ad data. This is to say you can see which specific actions an individual shopper took on which ads as they progressed through the funnel.
We can see what ads shoppers viewed, clicked, and converted against and how that customer was influenced by the various components of our advertising presence across platforms on their path to conversion.
Build multitouch or fractional attribution models
AMC helps determine which efforts are leading to incrementality and netting new customers, and how much credit they deserve for doing so. Consider the traditional method for attribution on Amazon which is last-touch. That means whatever ad campaign was the last one the shopper interacted with before their purchase is what hogs all the credit. Using AMC you can determine what other ads they interacted with on their path to purchase and weigh credit for each of those interactions differently. This gives a more “fair” distribution of credit.
A good example of this is video campaigns running both on and off Amazon through Amazon DSP. These campaigns in themselves don’t often earn conversions but you can now tie the data together to understand what impact they have on campaigns further down the funnel that are more performance-oriented. This allows you to then adjust those campaigns towards that objective.
Create clear assessment of the path-to-conversion journey
Every customer’s journey is unique, meaning it’s advantageous for advertisers who are able to analyze different audience interaction types, sequences, and frequencies. AMC also gives insights into engaged audience attributes and the composition of ad-exposed audiences.
An example here could be how long it takes a shopper to make a purchase and what ad touchpoints they have in what order with what time between them.
Analyze media mix
AMC offers expert analysis for helping brands and advertisers understand their channels’ different media combinations and each media channel’s incremental value. Through custom attribution, the contribution of different campaigns across different media channels and ad formats can be determined. This can help you understand how different media types mixed together to create the most potent mix.
Create custom, curated reports
AMC helps advertisers create reports that match their audience, tendencies, messaging, goals, channels, etc. for every campaign. These reports will have more details on revenue, impressions, and engagement than those in any other Amazon advertising platform.
Create specialized campaigns (in the near future)
Through AMC’s cross-channel reporting capabilities, data transparency is made possible, making the creation of specialized campaigns easier.
Manage audiences from inside AMC will come in Q1 2023. This will allow you to create an Amazon DSP audience via SQL right inside your AMC instance.
READ MORE: (Seller Sessions Podcast) The benefits of Amazon Marketing Cloud and how to get the most out of it
What can you NOT do with Amazon Marketing Cloud?
Like all analytical tools, there are limitations to AMC. So what can’t you do?
❌ Identify individual shoppers – AMC’s outputs are all anonymous to comply with Amazon’s privacy policies
❌ See data from all sponsored ad types – Sponsored Brands data is not yet included but is on the roadmap for next year (Sponsored Display was added in late June 2022).
❌ Use a simple point-and-click graphic interface to extract insights – AMC isn’t a visual interface where you can access each insight, and doesn’t have visualizations like charts and graphs built in.
❌ Pull reports from a predefined set like in the Advertising Console – Users will need familiarity with SQL and structured data at least, and possibly APIs, JSON, AWS services, and the use of other technical analytics tools like business intelligence software.
❌ See advertising data in real-time and make adjustments on the fly – AMC is not designed for this use case, what you are looking for is Amazon Marketing Stream. However, the data from AMS is part of what can be examined in AMC.
❌ Measure the on-Amazon impact of your off-Amazon ad spend but Amazon Attribution will help you accomplish this.
Who can use Amazon Marketing Cloud?
- Eligible advertisers need to have Amazon DSP campaigns live in the last 28 days, execute Amazon DSP Master Service Agreement (MSA) and they must also have a technical source who’s familiar with SQL to create queries. Next year Amazon may open AMC to all advertisers not just those spending with Amazon DSP.
- Agencies can also access AMC themselves as long as they have an Amazon Web Services (AWS) account or partner with a software tool that provides that function.
Both parties can use Amazon Marketing Cloud for free via web-based UI and API or take things to the next level by integrating other Amazon data with AMC data and analyzing it and building visualizations with a Business Intelligence software such as Zoho.
Advertisers from countries in North America (Canada, Mexico, US), South America (Brazil), Europe (Germany, Spain, France, Italy, The Netherlands, and the United Kingdom), Middle East (Kingdom of Saudi Arabia and the United Arab Emirates), and Asia Pacific (Australia, India, Japan, and Singapore) can benefit from AMC.
How can you access Amazon Marketing Cloud?
Amazon Marketing Cloud can be most easily accessed directly from a simple UI where you write SQL (Structured Query Language, a way of interrogating a database) queries and get data returned in the form of tables, or through an API. AMC has a growing library of suggested SQL queries to help programmers take advantage of the vast data sets inside the tool.
We at Pathfinder work with AMC through a software partner who helps us build visualizations we assemble in partnership with them.
To start, you need to have it activated by someone at Amazon. As of this writing, anything with AMC is very manual and having a point of contact with whom you can communicate is your best option. Another way is by contacting the software provider you’re working with to see if they can get in touch with someone at Amazon to help you with AMC access.
Another thing to note is that AMC instances are per-advertiser. It’s not like Amazon DSP where you might as an agency have multiple businesses nested in your umbrella Amazon DSP account. Instead, you’ll have access to multiple AMC instances but each of them is siloed off from each other and unable to interact.
✋➝ Amazon Advertising Management (PPC, DSP)
🗨️➝ Follow us on LinkedIn
📧➝ Subscribe to our newsletter
🛍️➝ We Design Amazon Stores
🔎➝ Amazon Advertising Audits
▶️➝ Subscribe to our YouTube channel