Amazon Marketing Cloud (AMC) has endless applications. Most brands either haven't started or aren't getting real value from it. Here are the diagnostic reports, audience builds, bid modifiers, and tests that consistently change how you spend, who you reach, and what shows up in revenue — drawn from Nectar's work across $100M+ in annual Amazon ad spend.
🎥 Watch the Full Session: Get the complete breakdown in the recorded webinar: AMC Levers That Actually Drive Revenue.
Amazon Marketing Cloud is Amazon's advertiser clean-room solution — a privacy-safe environment that combines signals from across Amazon's properties (Amazon Ads, Prime Video, Fire TV, Twitch, Amazon Fresh, Amazon Music, and more) with your first-party data. The output is three capabilities: audiences you can activate, measurement you can trust, and insights you can't get anywhere else, according to Amazon Ads.
The clean-room structure matters. Amazon anonymizes customer-level data so you can analyze cohorts — shoppers who added to cart, past purchasers, Subscribe & Save subscribers — without identifying individual buyers. That's what "privacy-safe" means in practice.
AMC is available to any brand running Sponsored Ads. DSP (Demand-Side Platform) is not required to access reports or build custom audiences. SQL knowledge is no longer required — Amazon's GenAI SQL assistant handles query generation from plain-English prompts.
Key Benchmarks from Nectar AMC Audits:
The AMC Activation Loop is a four-step operational framework for converting AMC data into revenue: Diagnose Revenue Drivers → Build Custom Audiences → Activate Ads & DSP → Measure & Repeat.
Most brands stall at step one. They run a path-to-purchase report, get surprised by the output, and move on. The loop only closes when the diagnosis triggers an action — a new audience, a bid adjustment, a campaign restructure — and then something is measured to confirm it changed the business.
The cadence is weekly, not quarterly. Some queries run every week (audience performance, frequency reports). Others have attribution windows requiring monthly review. The principle holds either way: data without action is noise.
Want to understand how your current AMC setup compares? Talk to our retail media team.
AMC has hundreds of possible queries. These six answer questions no other Amazon reporting tool can answer — and each closes directly into a strategic action.
Which campaigns and search terms drive genuinely new customers, and at what cost? A campaign that looks average on ROAS might be your best NTB driver. You'd never know without this report. The reverse is also true: a high-ROAS campaign might just be converting customers who already knew you. NTB rate and NTB cost-per-acquisition belong in every weekly review, alongside ROAS.
Which product is your brand's front door — where do new customers start before buying anything else? This changes where your ad dollar goes. One brand Harriet worked with was convinced their $24 hero product was the entry point. AMC showed a $14 companion item was actually driving the majority of NTB customers. The entire media strategy shifted once the true gateway was identified.
What do customers buy together, and in what sequence? This is the foundation of any cross-sell or bundle strategy. A lower-priced entry product is often the strongest predictor of a much larger second purchase. Knowing the sequence lets you structure retargeting windows around the actual customer journey, not assumptions.
The most discussed AMC report — and the most misread. In a real-account path-to-purchase analysis from this session, the top five conversion paths broke down as follows:
| Conversion Path | Sales | NTB Sales | Path Sales % |
|---|---|---|---|
| Sponsored Products × 2 → Purchase | $3,754,743 | $3,234,503 | 74.45% |
| Display (Sponsored Brands) → Purchase | $481,686 | $412,773 | 9.55% |
| Online Video (SBV) → Purchase | $205,729 | $177,177 | 4.08% |
| Display SB + SP × 2 → Purchase | $172,281 | $143,712 | 3.42% |
| SP × 3 + Display SB → Purchase | $130,269 | $104,221 | 2.58% |
Source: Amazon Marketing Cloud (AMC) retail insights, Nectar account audit
The takeaway most brands resist: Sponsored Products single-touch dominates. Multi-format sequencing is real but a small fraction of total volume. The implication isn't to abandon upper-funnel — it's to hold those formats to standalone performance standards, and use AMC's multi-touch attribution to understand incremental contribution rather than defaulting to last-click logic.
"The data shows that most customers just found them in search, clicked a Sponsored Product ad, and bought. That's not a failure of the funnel — that's just sometimes how customers behave." — Harriet Carson, VP of Advertising at Nectar
Which customers are worth the most over time? The AMC cohort retention table — each row a month of acquisition, each column the percentage of that cohort who returned in months 1–13 — tells a story most brands haven't seen clearly before.
What Nectar consistently sees: retention is strongest in months 1–3 (3–6% per month), declining to 2–3% by month nine. The first 60 days are where lifecycle marketing has the highest impact. In one 2025 audit, mid-2024 cohorts were outperforming early 2025 cohorts on early retention — a signal that something changed in acquisition quality or category competition. Finding out why those cohorts performed better is more valuable than almost any campaign-level optimization.
"The first 60 days are where you have the most influence over whether a customer becomes a repeat buyer." — Harriet Carson, VP of Advertising at Nectar
Action items: lifecycle marketing in the first 30 days, replenishment reminders and cross-sell in the first 60, reactivation campaigns for M6+ lapsed buyers.
Subscribe & Save (S&S) is one of the highest-value customer behaviors on Amazon — and most brands have almost no visibility into what drives enrollment or what causes churn. AMC's paid subscriptions change that.
In one audit using AMC retail insights data: 22% of S&S subscribers dropped off after their first delivery, 29% after the second. The practical activation: build an audience of customers who received a product via S&S four to 13 months ago but haven't received one in the past three months — then retarget them via DSP or Sponsored Display. The list is small, the CPCs are low, and the revenue recapture is material.
Two AMC paid subscriptions are worth activating: Amazon Insights and Amazon Retail Purchases. A 60-day free trial is available directly through Amazon Marketing Cloud.
What they unlock: organic purchase data (not just ad-exposed customers), five-year customer history vs. the default 365-day NTB window, and audience builds off organic purchase signals. For Vendor Central brands — which previously had almost no visibility into repeat purchase behavior, LTV, or customer journey data — this is a significant shift.
Pro tip: Don't activate the 60-day trial until you have specific use cases mapped out. Know which queries you'll run, what audiences you'll build, and what changes you'll track before day one.
Ready to run your first AMC cohort analysis or activate Paid Subscriptions? Talk to our Amazon team.
Diagnosis is only half the loop. This is where most accounts leave revenue behind — they have the data and never activate on it.
AMC audiences deploy via DSP and Sponsored Display (direct targeting) or as bid modifiers on Sponsored Products and Sponsored Brands. They organize into three buckets:
Reach Existing Customers — lapsed buyers, cart abandoners, S&S subscribers in a specific repurchase window, brand store visitors who didn't purchase. The most underused play: lapsed customers. Someone who bought eight months ago and hasn't returned isn't lost — they just need a reason to come back. A targeted DSP campaign with creative that acknowledges the prior purchase is consistently among the highest-ROI programs in mature accounts.
Reach Lookalike Customers — AMC builds lookalike audiences from your Amazon buyer base or DTC customer list: people who behave like your best buyers but haven't purchased yet.
Reach New Customers — True NTB suppression logic excluding anyone who purchased in the past five years. Cleaner new-customer targeting than what standard Amazon audiences offer.
New in 2026: AMC now includes a dataset allowing you to target shoppers who have a competitor's product in their cart. Early testing is showing meaningful results for category conquesting at the consideration stage.
The old model: bid modifiers were placement-based — top of search, rest of search, product pages. AMC changes the logic. Bid by who is searching, not just where the ad appears.
| Model | Bid Logic |
|---|---|
| Before AMC | Bid modifiers based on placement (top of search, rest of search, product pages) |
| With AMC | Bid by audience attribute. Placement is secondary. The searcher drives the bid. |
Case in point — premium eyewear bidding on "Glasses":
| Searcher Type | Willing Bid |
|---|---|
| General "Glasses" searcher | $0.50 |
| High-intent, premium, or NTB shopper | $3.00 |
Same keyword. 6× the bid. The audience attribute changes the value of the click. For premium brands that have historically struggled to compete on high-volume keywords without burning through budget, audience bid modifiers are the structural solution.
AMC audiences don't have to be used in isolation. Three activation patterns that consistently perform:
Sequence targeting: Target lookalikes of multi-ASIN purchasers via Sponsored Display → follow up with a bid boost in Sponsored Products for the segment that clicked but didn't convert.
Dual signal: Boost bids in Sponsored Brands for audiences who visited your PDP (Product Detail Page) and belong to a relevant interest segment → remarket with Sponsored Display to add-to-cart abandoners.
Same-audience simultaneous: Boost bids in Sponsored Products and Sponsored Brands, and target the same lookalike via Sponsored Display concurrently — for high-LTV shoppers who've spent $500+ at your brand across ASINs in the past 12 months.
AMC's tentpole event phase overlap report maps three custom date windows — Lead-In, Day-Of, and Post-Event — and shows how marketing activations across those phases interact to drive conversion. It's one of the most underleveraged tools in the platform.
| Phase | Window | What to Track |
|---|---|---|
| Lead-In | 2 weeks before | When first brand exposures happen — most occur here, not day-of |
| Day-Of | Event days (e.g., four days for Prime Day) | When ad-exposed users actually convert |
| Post-Event | 2 weeks after | Delayed conversions and new-segment buyers |
The most underleveraged phase: Post-Event. Category search surges on Prime Day even for shoppers who don't convert. Those audiences — PDP viewers, review page readers, add-to-wishlist users — are addressable for two weeks after. Save those lists. If they don't convert in the post-event window, retarget them at the next deal event.
Extension worth testing: Identify shoppers who had never been exposed to your brand before the event and purchased day-of. That's a different — and arguably more valuable — segment than deal-waiters, and a strong seed for lookalike expansion.
Building your AMC strategy for Prime Day 2026? Talk to our Amazon team.
If you looked at AMC a year ago and decided "not yet" — it's worth a second look. Four changes that matter in 2026:
GenAI SQL Assistant — Plain-English prompts now generate AMC-tailored SQL queries. Query development that used to take 30+ minutes now takes minutes. The barrier of SQL expertise is largely gone.
Paid Subscriptions — Amazon Insights and Amazon Retail Purchases are live and available with a 60-day free trial, unlocking 13-month NTB visibility and five years of retail transaction data.
Advanced Audience Templates — Amazon is releasing new audience and measurement templates frequently — check the template library every week. Many brands are months behind on activation simply because they haven't looked.
Expanded S&S Signals — Subscribe & Save audience plays that previously required custom workarounds are now first-class in AMC. If you have a meaningful consumables business and aren't using this, there's no longer a reason to wait.
Most AMC measurement is descriptive — it tells you what happened. These five tests are interventional — they show what's causing outcomes. Don't run all five at once. Pick the one with the biggest unknown. Three of the five require holdout groups, which need to be built into the test design before launch.
1. Multi-Touch Funnel Sequencing — Does sequential exposure (DSP video → Sponsored Brands Video → Sponsored Products retargeting) outperform single-touch? Is the added complexity worth it?
2. Purchase Cycle Recency — Compare 1-day, 7-day, 14-day, and 30-day retargeting windows. Where does retargeting stop adding incremental value?
3. Creative Variant — Lifestyle vs. product-first vs. UGC video across Sponsored Brands Video, Sponsored Display video, and DSP video. Which format earns the most impressions per dollar in your category?
4. Category Conquesting — Test conquesting in high-volume subcategories vs. a holdout. Is competitor targeting driving real NTB lift — or just impressions?
5. Upper-Funnel DSP Prospecting — Measure new-household reach from DSP prospecting. Quantify the brand-introduction lift you're paying for before scaling the budget.
Amazon Marketing Cloud is Amazon's advertiser clean-room solution — a privacy-safe environment that combines signals from across Amazon's properties with your first-party data to enable custom audience creation, cross-funnel attribution, and advanced measurement. It's available to any brand running Sponsored Ads and no longer requires SQL coding for templated queries.
Start with the New-to-Brand report and Path-to-Purchase report — they answer the two highest-impact questions: which campaigns drive genuinely new customers, and how shoppers actually move toward purchase. Add the Gateway Product report to understand where new customers enter your catalog. These three together reshape most brands' media allocation faster than any other starting point.
AMC lets you bid based on who is searching, not just where the ad appears. You set a base bid for any searcher, then apply a multiplier for high-value audience segments — NTB shoppers, repeat buyers, or specific behavioral cohorts. In practice, the same keyword can carry a $0.50 base bid and a $3.00 effective bid for a high-intent, NTB shopper. This is the primary activation method for Sponsored Products and Sponsored Brands.
No. AMC is available to any advertiser running Sponsored Ads, and DSP is not required to access reports, run queries, or build custom audiences. DSP is required to deploy those audiences via programmatic display targeting — but bid modifiers on Sponsored Products and Sponsored Brands work without it.
Amazon Insights and Amazon Retail Purchases include a 60-day free trial and are available through Amazon Marketing Cloud. They unlock organic purchase data, five-year customer history, and audience builds from non-ad-exposed shoppers. Vendor Central brands benefit most, gaining visibility into repeat purchase behavior and LTV that was previously unavailable. Plan your use cases before activating the trial — 60 days passes quickly.
AMC retail insights data shows S&S subscriber drop-off by delivery number — so you can see exactly when customers churn. Build a retargeting audience of lapsed S&S subscribers — customers who received a delivery four to 13 months ago but not in the past three months — and re-engage them via DSP or Sponsored Display. The list is small, the CPCs are low, and the recapture rate is typically high.
AMC custom audiences are built in the AMC console using templated or custom queries. They're available in your advertising campaigns within 48 hours. Deploy them for direct targeting in Sponsored Display and DSP, or as bid modifiers in Sponsored Products and Sponsored Brands. The three core buckets: existing customers (lapsed, cart abandoners, S&S subscribers), lookalikes (built from your Amazon buyer or DTC lists), and NTB audiences (excluding the past five years of purchasers).
Start with the New-to-Brand and Path-to-Purchase reports — they require no paid subscriptions and answer questions that reshape media strategy fastest. Once you have a diagnosis, build one audience (lapsed customers or cart abandoners are the fastest to ROI-positive), deploy it via Sponsored Display, and measure for 30 days. That first closed loop teaches you more than any report run in isolation.
The brands getting real value from Amazon Marketing Cloud aren't running more reports — they're closing the loop between insight and action faster than their competitors. Diagnose what's working. Build the audience it implies. Activate across Sponsored Ads and DSP. Measure whether it changed the business. If you want a second set of eyes on where your AMC setup is leaving revenue on the table, our team runs these analyses across accounts of every size.
Jason Landro is Co-Founder & Co-CEO of Nectar, where he leads strategy across the agency's Amazon and marketplace programs. He has spent more than a decade scaling 7- and 8-figure brands on Amazon, with a focus on shifting account economics from vanity ROAS to profitable, incremental growth. Connect on LinkedIn.