
Marketplace
The beginner's guide to Amazon Marketing Cloud (AMC) in 2026




Written by Darkroom leardership
12 min read
October 14, 2026
Amazon Marketing Cloud (AMC) is Amazon's no-cost clean room, built on AWS Clean Rooms, where advertisers query pseudonymized Amazon Ads signals alongside their own first-party data. It returns only aggregated results, and the audiences it builds can be activated in Amazon DSP and sponsored ads, so measurement and targeting run on the same data.
In 2021, Amazon launched Amazon Marketing Cloud in beta as a measurement tool for a small group of Amazon DSP advertisers.
Since September 2025, any brand running sponsored ads can open Amazon AMC straight from its Amazon Ads console, at no cost.
Here's everything you need to know about Amazon Marketing Cloud: what it holds, the four analyses that move budget, how to activate its audiences, and what it takes to run.
Key takeaways
AMC is free to use, so the real investment is the analyst time it takes to turn queries into budget decisions.
New-to-brand, path to purchase, media overlap, and tentpole lift are the four analyses to run first, because each one ends in a budget decision.
A path-to-purchase result shows what buyers saw before they bought, so large budget moves still need a holdout test to prove the ad caused the sale.
Audiences built in AMC can be activated in Amazon DSP, in Sponsored Display and as bid boosts in Sponsored Products and Sponsored Brands.
Results come back blank when a segment is too small, so start every new query on a wide date range.

What is Amazon Marketing Cloud?
Think of AMC as your own private analytics workspace, set up just for your business. It lets you dig into hundreds of different data points about your campaigns and then gives you back summarized results, so you can see the bigger picture without exposing any raw data.
Amazon calls AMC a clean room for analytics and audience building. What does that mean? Basically, it combines data from Amazon Ads and your own sources in a privacy-safe way using pseudonymized signals. You can either write your own queries or use ready-made templates to analyze things like impressions, clicks, and conversions.
What data lives inside AMC?
The types of questions you can ask in AMC really depend on the data available in your instance. As of October 2026, AMC supports five main types of data.
Amazon Ads campaign signals: ad-attributed impressions, clicks, and conversions from your sponsored ads and Amazon DSP campaigns, loaded automatically. Since January 2023, they include Sponsored Brands keywords, search terms, and placements, so you can study search behavior next to display and video exposure.
Your first-party signals: pseudonymized CRM records, site events, or offline sales that you upload or stream through Ads Data Manager, the AMC APIs, or a CDP partner.
Amazon Insights (beta): wider shopping and audience segment insights, plus third-party feeds such as Experian vehicle purchase data and NCS offline sales data.
Prime Video viewership signals: These include details like content title, content type, and show-level viewing data. This data has been in open beta since November 2025, and it can be especially useful for brands looking to connect media exposure with shopping behavior.
Household fields: household ID and household size, added on 6 May 2026 to measure reach and frequency across shared screens.
Most teams begin by exploring campaign signals, then layer in their own first-party data as they start to answer their initial questions. This step-by-step approach helps teams build a more complete picture over time.

How AMC differs from the reports in your Amazon Ads console
Think of AMC as your custom analytics toolkit. It lets you ask your own questions about your campaigns, no matter which ad types you are running.
Standard campaign reports are great for checking the usual metrics for each campaign and ad type. They work well for everyday bidding and optimization. But as soon as you want to understand how different ad types work together, those reports hit a wall.
For example, say you want to know how many shoppers watched your Sponsored Brands video in the first week and then came back to buy through a Sponsored Products click two weeks later. That kind of cross-ad insight is something only AMC can provide.
Amazon describes this difference similarly. According to its product page, AMC lets you dig into event-level data and create custom metrics while keeping privacy safeguards in place. Standard reports just can't offer that level of flexibility.
How does AMC work with AWS Clean Rooms?
AMC is built right on top of AWS Clean Rooms, but everything is managed through your Amazon Ads account. There’s no need to juggle multiple platforms or logins.
AWS Clean Rooms is Amazon’s solution for companies that want to analyze shared data sets without ever exchanging raw data. AMC uses this same technology as its foundation.
Amazon has taken this clean room technology and packaged it as part of the Amazon Ads suite. The best part is that AMC's clean room is ready to go out of the box, so you don’t need to set up or purchase anything extra in AWS.
What goes into your AMC instance
Each advertiser gets their own dedicated AMC clean room instance, which already includes all the signals from their Amazon Ads campaigns. This means there’s no need to manually import your campaign data.
If you want to bring in your own data, it needs to be pseudonymized before uploading. Once your data is inside your AMC instance, it stays there. According to Amazon, you can’t export it or let anyone else access it, including Amazon.
What comes back out
When you run a query, AMC only returns aggregated and anonymous results. You’ll see totals and rates for groups of shoppers, but never any information about individual people.
This setup is what makes it possible to safely combine your own customer data with Amazon’s signals. All the analysis happens inside the clean room, and only the summarized results ever leave it.
The types of analysis you can create with Amazon Marketing Cloud
There are four core types of analysis to focus on: new-to-brand, path to purchase, media overlap, and tentpole lift. Each one leads to a clear decision you can put into action this quarter.
Amazon Marketing Cloud can answer hundreds of questions, but only a few will actually shift where your budget goes. It makes sense to start with these four analyses, run them regularly, and only add more once each one has helped you adjust a budget line.
Amazon’s instructional query library, available since October 2021, offers ready-made queries for common scenarios like path to conversion. Each query comes with helpful notes so you can understand what the results mean and how to use them.

New-to-brand analysis
The key decision here is to shift budget toward the campaigns and keywords that are bringing in first-time buyers at a cost that fits your goals.
Amazon considers an order new-to-brand if the shopper hasn’t purchased from your brand in the last 12 months. While the console shows this at the campaign level for certain ad types, AMC lets you break it down further by ad type, keyword, ASIN, and even the sequence of ads that led to the order.
It’s a good idea to check the new-to-brand share for each campaign and top keyword. Usually, this reveals two groups: some campaigns are great at bringing in first-time buyers, while others mostly drive repeat purchases from existing customers who probably would have bought again regardless.
For the campaigns that mostly reach existing customers, it makes sense to cap spend based on your profit targets. The campaigns that attract new customers deserve more budget, as long as the cost per new customer fits within your payback period.
Path to purchase
The main takeaway is to protect the ad formats that consistently appear early in the customer journey, even if last-click reports make them look less valuable.
The path-to-conversion query shows which combinations of ad types shoppers saw before purchasing, along with the number of purchases and sales for each path. This is where upper-funnel formats like Sponsored Brands video, Streaming TV, and display ads finally get the recognition they deserve.
If an ad format shows up early in a lot of high-value customer journeys, cutting it just because its last-click ACoS looks high can quietly reduce the number of conversions down the line.
One important caveat: a path only shows what buyers saw, not what actually caused the sale. Before shifting a large portion of your budget based on a path analysis, it’s smart to confirm the finding with an incrementality test that holds a group back from seeing the ads.
Media overlap
The decision here is to invest in ad-type combinations that actually increase purchase rates, and to limit frequency where extra exposure doesn’t add value.
Overlap analysis compares shoppers who saw just one ad product with those who saw combinations of Sponsored Products, Sponsored Brands, Sponsored Display, and Amazon DSP. Amazon’s AMC examples show how to measure both the total reach and the incremental reach of each media type.
It helps to look at this analysis from both angles. If shoppers who see two ad types buy at a much higher rate than those who see just one, the combination is worth the investment. If the purchase rate barely changes, it means you’re likely paying twice to reach the same shopper.
The new household-level fields coming in May 2026 make this even clearer. Measuring reach and frequency per household helps spot over-exposure on shared devices like TVs and tablets, which device-level counts can miss.
Tentpole lift
The decision here is to use last year’s data to set the lead-in start date and lead-out budget for your next big event.
For events like Prime Day, Prime Big Deal Days, or Black Friday, it helps to break the event into three windows: lead-in, event, and lead-out. Then, ask two key questions: When did buyers first see your ads, and what was the new-to-brand rate in each window?
If most event buyers first saw your ads just a week before the event, starting your lead-in three weeks early means you’re paying for impressions that probably won’t convert. On the other hand, if the lead-out window keeps converting shoppers who first saw your ads during the event, cutting spend right after the event ends could mean missing out on extra sales.
It’s a good idea to run the same analysis on last year’s event and this year’s, side by side, before finalizing your next plan.
Read also: Ecommerce Analytics: The Attribution and Measurement Stack That Actually Informs Decisions
How do you activate AMC audiences in Amazon DSP and sponsored ads?
Build a rule-based or lookalike audience in AMC, set how often it refreshes, and it appears in your DSP and sponsored ads accounts ready to use.
That's what turns AMC into a targeting tool as well as a reporting one. The same query that finds a segment can turn it into an audience. If you also buy Amazon DSP, those audiences show up there automatically.
Here's where Amazon Marketing Cloud audiences can be activated today:
Destination | How the AMC audience is used | Available since |
|---|---|---|
Amazon DSP | Target or exclude on a line item, like any other DSP audience | May 2023 |
Sponsored Display | Audience targeting | October 2024 |
Sponsored Products | Audience bid boosting, with its own performance report in the console | API October 2024, console November 2024 |
Sponsored Brands | Audience bid boosting | October 2024 |
DSP activation launched with AMC Audiences in May 2023, and Amazon extended AMC audiences to Sponsored Display targeting and bid boosts in October 2024. Sponsored Products bid boosting reached the Amazon Ads console on 4 November 2024.
Which audiences are worth building? Amazon's own launch examples give you four strong starting points:
Cart abandoners: shoppers who added your product to cart after seeing an ad and didn't buy.
Streaming-only viewers: shoppers who saw your Streaming TV ads but no display ads yet.
Lapsed buyers: customers who bought last year and haven't engaged recently.
Over-exposed shoppers: people who've seen your ads across channels many times, used as an exclusion.
For bid boosts, set the boost so the boosted audience still lands under your break-even ACoS. A boost that pushes a warm audience past break-even just buys orders at a loss faster.
Amazon's own advice is to start with standard Amazon audiences, then use AMC for what they can't do, such as multi-channel exposure rules or your own segment definitions. AMC audiences also follow the same size and policy rules as any other Amazon audience, so very narrow segments won't activate.
The upside is real when the audience is right. In an Amazon case study, partner Buy Box Experts built AMC audiences for SimpliSafe and reported ROAS up 109%, cost per unit sold down 65% and a 29% lift in new-to-brand share of purchases (partner-provided data, US, 2023).
Teams running this at scale can automate refreshes and activation through the Amazon Marketing Cloud API.
Step-by-step to create and activate an AMC audience
Here's the path from an empty AMC instance to a live audience in your campaigns:
1. Open AMC: in the Amazon Ads console, go to the left-hand navigation, open Measurement & Reporting, and select Amazon Marketing Cloud.
2. Go to the Audiences tab: this is where every audience in your instance is created and listed, as Amazon's lookalike audience launch shows.
3. Choose how to build it: start from a template in the use case library, use a no-code option such as the high-value audiences solution in the Solutions tab, describe the audience to Ads Agent through "Ask AI for AMC help", or write your own SQL in the query editor.
4. Pick the audience type: rule-based for remarketing and exclusions, or lookalike to find new shoppers who resemble a seed group, such as your best customers.
5. Set the refresh frequency: daily refreshes keep cart abandoners and recent viewers current, while slower-moving segments like lapsed buyers can refresh less often.
6. Run it and check the size: AMC audiences follow the same size and policy rules as other Amazon audiences, so widen the lookback window if yours comes back too small.
7. Activate it in Amazon DSP: the audience appears in your DSP account automatically, and you select it on a line item like any other audience.
8. Activate it in Sponsored Products: in a new or existing campaign, open the Audience Bid Adjustment tab and add a bid multiplier for the AMC audience.
9. Activate it in Sponsored Display and Sponsored Brands: target the audience directly in a Sponsored Display campaign, and add bid boosts for Sponsored Brands through the Amazon Ads API or a tool built on it.
10. Track the result: the bid-boosted audience report for Sponsored Products sits under the Campaigns tab in the Ads console.
Teams running this at scale can automate the same refreshes and activations through the Amazon Marketing Cloud API.
Who can access Amazon Marketing Cloud in 2026?
Any brand running sponsored ads can open AMC from its own Amazon Ads account, and DSP advertisers and partners have their own routes in.
Since 16 September 2025, sponsored ads advertisers have had self-service access inside their Amazon Ads account. You'll find it in the left-hand navigation under Measurement & Reporting.
According to Marketing Dive's report on the launch, it covers anyone running Sponsored Products, Sponsored Brands, Sponsored Display, or Sponsored TV, and it ships with no-code templates.
That was the last step in a two-year opening. AMC started in 2021 as a tool for DSP advertisers. On 15 October 2024, Amazon dropped the DSP requirement for brands working through registered partners. The September 2025 launch removed the partner step.
Two other routes still exist. Advertisers using Display, Video, and Audio with advanced settings request access from their Ad Tech Account Executive, and any brand can run AMC through a partner registered on the Amazon Ads Partner Network.
If you already have an instance, the AMC login sits in that same Measurement & Reporting menu.

Source: Ecom Brainly
Minimum requirements to use AMC
Amazon has four main requirements for using AMC, and interestingly, it doesn't list a minimum ad spend.
Before getting started with AMC, you need a few things in place.
First, an active Sponsored Ads account is needed. AMC is available to any advertiser running Sponsored Products, Sponsored Brands, Sponsored Display, or Sponsored TV campaigns.
AMC is only available in certain marketplaces. It's currently live in 20 regions, including the US, Canada, Mexico, Brazil, the UK, Germany, Japan, and India.
The way to access AMC depends on your setup. Sponsored ads advertisers can use the self-service option in the console. If you are running Display, Video, or Audio campaigns with advanced settings, access is requested through your Ad Tech Account Executive. Brands can also work with a partner from the Amazon Ads Partner Network.
If you plan to bring your own customer data, make sure it is pseudonymized. AMC only accepts inputs that protect customer identities.
Once these requirements are met, you can get started with your own AMC instance. The value you get from AMC will depend on factors like your order volume and the mix of media you use, which is something to consider next.
When is a brand ready to put AMC to work?
When you have the volume, the media mix, and an owner. Without all three, AMC produces blank tables and unread reports.
Use this checklist before you staff it:
Volume: enough monthly orders that your segments clear the aggregation thresholds. Test with a 90-day query before planning anything.
Mix: at least two ad types running. With one, path-to-purchase and overlap analyses have nothing to compare.
Questions: each standing query is tied to a named budget decision, such as lead-in timing or new-to-brand spend caps.
Owner: one person runs the queries on a fixed cadence and owns the changes that follow.
Truth test: a holdout plan for the conclusions that would move the most money.
If you tick four or five, start with the new-to-brand analysis this month. It's the fastest one to turn into a budget change.
If you tick fewer, use the no-code templates to build the habit while you fix the gaps. A simple query run every month teaches your team the most.
Read also: Amazon Marketing Strategies That Actually Work in 2026
How can advertisers use AMC wisely?
Think of AMC as your go-to decision tool when it comes to managing your Amazon budget. Here are six habits that have proven effective for enterprise brands, and they can work just as well for any team looking to get more out of AMC.

1. Start every query with a budget decision
Start by writing down the decision you want to make before diving into the query itself. For example, asking whether Sponsored Brands video should keep its share of the budget next quarter is exactly the kind of question AMC can help answer. On the other hand, broad questions like 'What do our customers look like?' might lead to a nice-looking chart, but often don't drive any real change.
If it's not clear which part of the budget a query could impact, it's usually best to set that question aside for now.
2. Run the same queries on a fixed calendar
Choose a handful of key queries to run on a regular schedule. For example, checking new-to-brand and overlap metrics each month, reviewing path to purchase every quarter, and running your main query before and after big events can help keep your analysis consistent and actionable.
Keeping a straightforward log for each run, including the date range, results, and the decisions made, can be a game changer. Over time, this log becomes a valuable benchmark for your brand, and it ensures continuity even as team members come and go.
3. Judge each ad type on the job it does
Sponsored Products are designed to close sales, so it makes sense to evaluate them based on ACoS and whether they hit your break-even point. In contrast, Sponsored Brands video, display, and Streaming TV are more about starting the customer journey. For these, focus on metrics like new-to-brand share and how frequently they appear early in the conversion path.
AMC provides that extra layer of data you need to evaluate upper-funnel formats fairly when it's time to set budgets.
4. Test the big calls before you move budget
AMC can show exactly what shoppers who saw your ads ended up doing. If you find something that could lead to a big budget shift, it's smart to run a holdout test first. That means pausing the format for one audience or region, keeping it active for a similar group, and then comparing the results.
For smaller decisions, AMC data alone is often enough. But for bigger moves, a test is worth the extra effort.
5. Turn every insight into an action within a week
One of the best things about AMC is that measurement and targeting use the same data. If a query uncovers a segment that's worth reaching, try building that audience right away and activating it in DSP or as a bid boost within the same week.
After about a month, check back in AMC to see how that audience performed. This cycle of moving from insight to audience to measurable results is where AMC really proves its value.
6. Bring your own customer data in early
Your CRM holds valuable information that Amazon doesn't have, like which customers buy directly from your site or have active subscriptions. Uploading this data, in a privacy-safe way, during your first month can make a big difference.
This approach lets you exclude current subscribers from acquisition campaigns, understand how your Amazon ads reach your top customers, and avoid spending money to win over shoppers you already have.
Turn AMC answers into Amazon budget decisions with Darkroom
An AMC insight only matters if it actually impacts your budget. With Darkroom, all your Amazon campaigns (Sponsored Products, Sponsored Brands, Sponsored Display, DSP, and AMC) work together as one system. That means when you spot a new-to-brand shopper or see a spike around a big event, those insights can quickly shape next week's bids and audience targeting.
Take Laundry Sauce as an example. By keeping Sponsored Ads and DSP campaigns in sync, they saw Amazon net revenue jump 290 percent and repeat order rates rise 23 percent.
It all starts with a 30-minute working session alongside a senior marketplace strategist, focused on your catalog and how to make your ads more efficient.
Book your free Amazon audit today.
Frequently asked questions
How much does AMC cost?
AMC is free. Amazon makes it available at no cost to eligible advertisers through its web interface and API, and sponsored ads advertisers can open it from their Amazon Ads console. The only paid option is Paid Features (beta), an optional subscription for extra shopping and third-party data, and Amazon doesn't publish its price.
Do you need SQL skills to use AMC?
Not to start. Amazon AMC includes no-code templates and an instructional query library for common analyses, and Ads Agent can write analytics and audience SQL from a plain-English prompt. You still want someone who can read SQL, because every template needs adjusting to your campaigns, date ranges and ASINs before its answer is trustworthy.
How is AMC different from Amazon Attribution?
Amazon Attribution measures how marketing outside Amazon, such as search, social, email and influencer campaigns, drives shopping on Amazon, using tags and a 14-day attribution window. AMC analyzes your Amazon Ads exposure and conversion signals in a clean room. Both are free, and brands sending off-Amazon traffic to their listings usually run both.
Can you upload your own customer data to AMC?
Yes. You can upload or stream pseudonymized first-party signals, such as CRM records or offline sales, through Ads Data Manager, the AMC APIs, or integrated partners like LiveRamp and Treasure Data. The data stays in your dedicated instance; Amazon can't export or access it, and every result comes back aggregated.
How far back does AMC data go?
Since November 2025, AMC keeps 25 months of ad traffic signals in the US and other early markets, up from 13 months, with more locales added in the first quarter of 2026. That window lets you compare the same tentpole event year over year and read seasonal patterns across two full buying cycles.
Is there an Amazon Marketing Cloud certification?
Yes. Amazon Ads Academy offers an AMC Certification built for people new to AMC or with little technical background. It runs across eight courses on measurement, custom audiences and first-party data. A separate beginner course covers Ads Agent in AMC, the AI assistant that writes queries and builds audience segments from natural language.
Can brands that don't sell on Amazon use AMC?
Yes, through Amazon DSP. Non-endemic advertisers can bring their own pseudonymized data and, since November 2025, analyze Prime Video viewership signals to see how their customers overlap with Prime Video audiences. UNICEF USA, advised by partner DELVE, used AMC insights to reach one-time donors and measured 366% direct ROAS (DELVE, US, 2023).

