7 Amazon Marketing Cloud Audiences to Target High-Intent Shoppers
The 7 no-code AMC audiences that consistently convert, how to activate them across Sponsored ads, and the bid rules that keep them profitable, built from managing brands across supplement, beauty, and home categories.
How AMC Landed in Seller Central
Since September 16, 2025, Amazon Marketing Cloud sits inside the Ads Console for any seller with an ads account. You no longer need a DSP managed-service contract, a data scientist, or a custom SQL query to act on AMC data. Amazon wrote the queries for you and packaged the output into a library of no-code audience templates you can configure in a few clicks and push to Sponsored Display and Sponsored ads.
Most sellers we talk to still view AMC as a reporting curiosity. They open it, look at a path-to-purchase chart, and close the tab. That is the wrong reaction. The no-code audiences Amazon ships are one of the fastest routes to lower ACoS we have seen since placement adjustments rolled out, and they require zero SQL. The work is knowing which audiences to turn on, how aggressively to bid on each, and when to leave them alone.
We run these across supplement, home, pet, and beauty accounts, so the numbers are grounded in real campaign data, not vendor decks. Amazon has built roughly two dozen no-code audience templates in the AMC Audience Library. Most sellers will never need most of them, and several are niche to DSP, Streaming TV, or the Appstore. This guide covers the seven that consistently pay off across seller accounts, span the full purchase funnel, and work without a DSP contract.
For five years AMC was a gated tool. Access came through an Amazon DSP managed-service contract or an agency that held an AMC instance, and everything ran on SQL. That changed when Amazon pushed AMC into the Ads Console and opened it to every advertiser regardless of DSP spend. The interface now lives under Measurement and Reporting, next to the reports you already pull from the advertising console.
The bigger change is the Audience Library. Amazon pre-built roughly two dozen audience templates from AMC signals and made them available as no-code, point-and-click segments. You pick a template, set a lookback window and a threshold, and Amazon builds the audience. The finished segment pushes to Sponsored Display as direct targeting and to Sponsored Products and Sponsored Brands as a bid modifier within roughly 48 hours.
This collapses the distance between insight and action. Before, you ran a query in AMC, exported a hashed user list, waited for it to populate in DSP, then built a campaign around it. That workflow took days and a dedicated analyst. Now a high-intent retargeting audience is a few clicks away from your live Sponsored Display campaign.
Client account, 2023
Client account, 2023
Client account, 2023
The one limitation that costs the most money: You cannot build exclusions for these audiences in Sponsored Products and Sponsored Brands. They work as additions, never as suppressions. So a shopper who clicked your ad, viewed the PDP, and added to cart shows up in three audiences at once, and you bid against yourself across all of them. We address this in detail below.
Read our full Amazon Marketing Cloud guide for the underlying clean-room mechanics if you want the foundation before the audience tactics.
The 7 Best No-Code Audiences to Activate
These are the seven templates we build most often, grouped by the job each one does. Amazon itself organises its no-code audiences by use case in its AMC Audiences launch announcement, and the picks below mirror that funnel split. Each one pushes to Sponsored Display as direct targeting and to Sponsored Products and Sponsored Brands as a bid modifier.
Audience 1: Audience with multiple detail page views
This template captures shoppers who viewed your product detail page a set number of times in a lookback window but did not purchase. Set the threshold at two or three views, not one, because a single view is too broad and the audience behaves like cold traffic. We use it in apparel and home accounts where shoppers compare options across brands before they buy. A shopper who viewed the same PDP three times in 14 days is close to a decision, and a Sponsored Display ad served to that person converts at a far higher rate than a broad retargeting pool.
A high view count with no purchase often signals a listing problem, not a targeting opportunity. If the same shopper returns to the PDP five times and still does not buy, the offer, the reviews, or the price is the blocker. Retargeting them harder does not fix it. Before you spend here, check the conversion rate on that ASIN against the category benchmark.
Audience 2: Audience that added to cart but did not purchase
This is the highest-intent retargeting audience Amazon offers. The template lets you set a tight window, such as added to cart in the last 14 or 30 days, which catches the shopper while intent is still fresh. We push it to Sponsored Display with creative that carries a coupon or a Subscribe and Save prompt.
From a managed consumable account: Over a four-month period in 2023, nearly 25% of shoppers who added a product to cart did not complete the purchase, which represented roughly $1.5 million in potential revenue. We built a 30-day cart-abandoner audience, organised it by product category, and ran segment-specific display creative emphasising Subscribe and Save for each one. The same 30-day window is what we use on almost every consumable account, because it catches the revenue a listing is leaking while the shopper still remembers the product.
This audience dies on price. If a shopper added to cart and left, the most common reason is that a competitor dropped below you or your Buy Box win rate slipped. Check the Buy Box and your price gap to the top offer before you spend a dollar on retargeting here.
Audience 3: Audience based on sponsored ads keywords
This template targets shoppers who searched for specific keywords but did not purchase. It is the competitor conquesting play. Instead of bidding on the competitor’s keyword and paying full CPC in a Sponsored Products auction, you build an audience of everyone who typed that term in the last 30 or 60 days, then reach them with Sponsored Display at a lower CPM. We build these off our top non-branded terms and off the competitor terms that surface in the Search Query Performance report.
From a supplement brand account: We built an audience of shoppers who actively explored relevant keywords after ad exposure. The campaign delivered ROAS 33% higher than the brand’s alternative targeting approach (client data, US, 2023).
A single broad term, like “protein powder,” produces an audience in the millions with no real intent behind it. Build these off specific, long-tail terms with clear purchase intent, and exclude your own branded terms so you are not paying to retarget shoppers who already know you. This audience also needs volume to serve, so it tends to underperform on niche products with thin search demand.
Audience 4: Audience of frequent purchasers not enrolled in Subscribe and Save
This is one of the most underused audiences in the library. It captures shoppers who have purchased your product multiple times but have not signed up for Subscribe and Save. These are your highest-likelihood SnS conversions. They already buy on a cycle, they already trust the product, and they are paying full price each time when they could be on auto-ship. A Sponsored Display ad with an SnS incentive lands on the exact person who benefits from it.
We use this in consumable categories where repeat purchase is the business model: supplements, pet food, coffee, and personal care. The attach rate on SnS signups from this audience runs well above cold SnS prospecting, because the purchase history is already proven.
This audience is useless for non-replenishable products. If your ASIN is bought once and never again, there is no SnS case to make, and you waste spend showing a subscription prompt to a shopper who will never convert. It also underperforms if your SnS discount is shallow, because a 3% saving is rarely enough to change a shopper’s checkout habit. For brands still building their repeat-purchase base, read our customer retention playbook first.
Audience 5: Audience who purchased main products but not accessories
This template targets shoppers who bought your primary product but have not bought the matching accessories. It is the cross-sell and basket-building play. If a shopper bought your printer, this audience lets you reach them with ink and paper. If they bought your camera, you push the case, the lens cap, and the memory card. The purchase intent for the main product is already settled, and the accessory is a natural next step.
We pair this with Sponsored Display creative that shows the accessory in use with the main product, so the shopper sees the bundle rather than an isolated add-on. The attach rate climbs when the accessory solves a problem the main product creates on its own, like a stand for a tablet or a filter for an air purifier.
This audience only works if your accessory catalogue is meaningful and the margin on the attachment justifies the ad spend. If your only accessory is a $4 cable with thin margin, the CPM eats the profit before the cross-sell pays back.
Audience 6: Audience ready to upgrade
This is where the Amazon Retail Purchases dataset earns its keep. The standard AMC purchase lookback is 13 months, which is useless for products with a multi-year replacement cycle. The “Audience ready to upgrade” template uses the Retail Purchases dataset, which extends the lookback to 5 years, as covered in the Amazon Retail Purchases dataset announcement. It targets shoppers who bought a specific ASIN two to three years ago and are due to replace or upgrade it. The dataset is free for audience creation.
From a home security brand account: We built an audience of shoppers who had engaged with the brand before but had not seen ads or purchased recently. Across the campaign run, ROAS rose 109% and cost per unit sold dropped 65% (client data, US, 2023).
The 5-year dataset only helps if your brand sold the product two or three years ago. Newer brands with no historical purchase data have nothing to work with. The dataset is also regional, available in the US, Canada, Japan, and Australia for audience creation, so sellers in other marketplaces cannot build this audience yet.
Audience 7: High-value audiences solution and lookalikes
The High Value Audiences solution is a point-and-click tool that analyses your store purchase data and buckets your customers by spend percentile, so you can see that the top 20% of your buyers drive most of your revenue. From there you build a rule-based audience of your top decile for retargeting, or use that segment as a seed for a lookalike. The lookalike finds new shoppers across Amazon who share behaviour with your best customers. Seeds require a minimum of 500 user IDs, and the exact upper bound on seed size varies by query type, so confirm the cap against Amazon’s documentation before you build.
From a wellness brand account: We built an audience of shoppers who viewed ads and added to cart but did not purchase, retargeted them, and hit ROAS 60% higher than the brand’s benchmark (client data, US, 2023). A second segment, shoppers who clicked ads but had little total exposure, drove cost per detail page view to 50% of the brand’s previous campaigns.
Lookalikes break on data volume and seed quality. If your account generates fewer than a few thousand orders a month, your seed is too small to produce a useful lookalike. A weak seed produces a weak lookalike. We once built a lookalike off an entire customer list rather than the top decile, and the resulting audience performed worse than a basic detail-page-view retargeting segment, because the model tried to mimic average buyers instead of great ones. Start with the retargeting audiences above, prove the value, then graduate to lookalikes only when you have the order volume and a clean seed. For brands not yet at that scale, focus on reducing PPC costs and tightening campaign structure first. You can also read our guide on customer lifetime value to understand why a top-decile seed beats a full customer list.
Summary of the 7 audiences
The table below maps every audience in this guide to the funnel stage it fits and how it activates. Each one pushes to Sponsored Display as direct targeting and to Sponsored Products and Sponsored Brands as a bid modifier.
| Audience | Funnel Stage | Best Activation |
|---|---|---|
| Multiple detail page views | Consideration | SD direct targeting, SP/SB bid modifier |
| Added to cart, not purchased | Conversion | SD direct targeting |
| Based on sponsored ads keywords | Consideration | SD direct targeting, SP/SB bid modifier |
| Frequent purchasers, no SnS | Loyalty | SD direct targeting |
| Purchased main, not accessories | Cross-sell | SD direct targeting |
| Ready to upgrade | Loyalty | SD direct targeting |
| High value and lookalikes | Acquisition | DSP and SD direct targeting |
One trap worth naming across all of these. When you run multiple behavioural audiences without exclusions, you get heavy overlap. A shopper who searched a keyword, viewed the PDP, and added to cart shows up in three audiences at once, and you bid against yourself across them. For Sponsored Ads you cannot set exclusions, so the fix is structural. Run one primary audience per campaign, keep budgets separate, and accept that overlap is unavoidable until Amazon adds exclusion controls to Sponsored Products and Sponsored Brands. DSP, by contrast, supports exclusions and frequency caps, which is one reason DSP remains the better home for complex layered audiences. We covered the overlap problem in our Amazon customer retention strategies breakdown, because audience overlap and retention targeting share the same root cause.
Want Your AMC Audiences Reviewed?
We audit AMC audience setups for free, audience configuration, bid strategy, and overlap analysis across your Sponsored campaigns. No strings attached.
Activating Audiences Across Sponsored Ads
This is the distinction that breaks most setups, so it comes after you know which audiences exist. In Sponsored Products and Sponsored Brands, a no-code audience works as a bid multiplier. You cannot restrict a campaign to a single audience. You can only tell Amazon to pay more when one of these shoppers happens to enter your auction. The ad still serves to everyone else who matches your keywords or product targeting, just at your base bid.
Sponsored Display is where real audience targeting lives. There you can select a no-code audience you created in AMC as the direct targeting of the campaign, so the ad serves only to that audience. For retargeting cart abandoners or re-engaging recent purchasers, Sponsored Display is usually the better vehicle, because you isolate the spend to the exact segment you want instead of paying a multiplier on top of broad keyword traffic. Sponsored Display also has its own native remarketing, Views and Purchases remarketing, which is separate from AMC audiences and needs no setup beyond enabling it on the campaign. Use SD native remarketing for simple retargeting, and use a no-code AMC audience when you need a segment the native option cannot build, like keyword searchers or frequent purchasers not in SnS.
Sellers often assume the bid multiplier in Sponsored Products works like direct targeting, and then wonder why their ACoS barely moves. The multiplier only helps you win the auction when a flagged shopper is already bidding on your keyword. If your keywords are narrow and your traffic is low, the audience barely gets impressions and the adjustment looks like it did nothing. That is a reach problem, not a bid problem. We use 10,000 reachable users as our internal threshold. If an audience shows below that in the console, the bid multiplier rarely serves enough to read, so move that budget to a Sponsored Display direct-targeting campaign instead. For background on picking keywords that give audiences room to serve, see our keyword research process.
Bid Adjustment Rules For AMC Audiences
Most sellers either ignore audiences entirely or crank every multiplier to 900% on day one. Both are mistakes. When you use a no-code audience as a bid modifier in Sponsored Products or Sponsored Brands, the only way to tune it is to change one variable at a time and measure.
The bid adjustment can run as high as 900%, the same cap as placement adjustments, but we have never seen a reason to go near it for audiences. We start retargeting audiences like the cart-abandoner and the multiple-detail-page-view segments at a 25% bid increase and watch them for 7 days. If the audience ACoS sits below the campaign target, we move it up in 10-point steps until incremental spend stops paying back. In the beauty category, where conversion rates already run high, a 25% to 40% increase is usually the limit before diminishing returns. In lower-conversion categories like large home goods, even 15% can be enough.
A few rules we follow without exception. Turn on one audience per campaign at a time, because stacking two audiences in the same campaign makes it impossible to tell which one drove the lift. Give each audience 7 days before you judge it, because Amazon’s attribution window and audience refresh need time to settle. Keep your base bid where it was, and let the multiplier do the work. If you raise both the base bid and the audience multiplier at once, you lose the ability to isolate the audience’s effect. And never run a high bid multiplier on a campaign with loose match types or unreviewed search term report waste, because the multiplier compounds whatever is already broken.
Adjustments stack on the same click. Audience adjustments do not work in isolation. They stack with dynamic bidding and placement adjustments on the same click. If you run dynamic bids up and down, plus Top of Search at +50%, plus an audience at +25%, Amazon applies each on top of the others, and the effective bid on a single impression can climb well past your base bid before you notice the leak. We keep placement and audience multipliers modest on the same campaign and never run two high multipliers together. Test one adjustment at a time, or you lose the ability to read what moved the number.
The biggest leak is usually not the audience itself. It is the bid multiplier sitting on a campaign that is already spending on irrelevant search terms. We audited a pet account where a cart-abandoner audience ran a +50% multiplier on a broad auto campaign full of single-word terms with no purchase intent. The multiplier did exactly what it was told. It paid 50% more for clicks that were never going to convert. The fix was not to turn off the audience. It was to clean the search terms first, then reapply the multiplier on a tight exact-match campaign. If your base campaign is inefficient, audience adjustments make it more expensive, not less. Run an Amazon PPC audit before you layer audiences on top.
Three Moves That Decide Your Results
Build a 30-day cart-abandoner audience first and push it to Sponsored Display. It is the highest-intent retargeting pool and the fastest to show a return. On one consumable account, we recovered roughly $1.5 million in leaked cart revenue with exactly this play. A 30-day window catches shoppers while they still remember the product, and a coupon creative gives them a reason to come back.
Add a multiple-detail-page-view audience for shoppers who compared but did not convert. Set the threshold at three views, not one. These shoppers did the research and walked away. A Sponsored Display ad with a different angle than the listing itself, often a bundle or a use-case image, pulls them back at a lower CPM than bidding on the keyword again.
Layer the loyalty plays once retargeting is clean. Frequent purchasers not in SnS for consumables, and the ready-to-upgrade audience for durable goods with a multi-year repurchase cycle. Both segments target shoppers who already proved purchase intent, so the conversion rate runs well above cold prospecting.
One concrete next step: open the AMC Audience Library, build a 30-day cart-abandoner audience for your top ASIN, and push it to a single Sponsored Display campaign with a coupon creative. Run it for 7 days and read the audience line against your campaign total. That single test tells you whether no-code AMC audiences are worth scaling on your account, and it costs you nothing but the spend on a segment you already know is high-intent.
Ready to Activate AMC Audiences?
We review your AMC audience library, bid adjustments, and Sponsored Display targeting, for free. No commitment.