Amazon Marketing Cloud (AMC) & Incrementality Analysis
Proving Advertising Incrementality and ROI
AMC gives you customer-level data across every ad touchpoint on Amazon. Most brands stop at dashboards. We use it to identify which campaigns drive new customers versus recapture existing demand, find where Sponsored Ads and DSP overlap is wasting budget, and build custom audiences you can't create anywhere else. The result isn't a better report. It's a different way to allocate spend.
The core question: are your ads creating new sales, or taxing demand that would have happened organically? We run holdout group analysis to quantify true advertising lift. We measure incremental ROAS instead of blended ROAS so you know which campaigns actually generate new demand. And we separate new-to-brand acquisition cost from reactivation cost so budget moves toward growth, not credit-taking.
AMC analysis answers questions. Custom audiences act on them. We build audiences from signals that only exist inside AMC: category entrants who bought a competitor but never tried your brand, NTB buyers from 90 days ago who never reordered, entry-product buyers who haven't discovered higher-margin lines, and multi-touch converters who engaged both SP and DSP. These audiences get pushed directly to DSP campaigns. That's where AMC stops being a reporting tool and starts moving revenue.
AMC Analysis Capabilities
FAQ
When you’ve outgrown pre-built reports, meaning your questions no longer match any canned dashboard. Helium 10 and similar tools are rules-based: they tell you what’s happening against known benchmarks. AMC is query-based: it answers questions no one’s pre-built for you ("what’s my path-to-purchase overlap between Sponsored Products and DSP for new-to-brand customers who came from competitor-conquest keywords?"). The trigger for graduating is usually (a) DSP spend above $50K/month, (b) cross-ad-type path analysis Sponsored Ads reports can’t produce, or (c) stitching Amazon behavior into a larger data model across DTC, retail-media networks, and CRM. Under those conditions, AMC typically cuts 15–25% of ad waste inside two quarters. Outside them — a $2M brand running Sponsored Products only — AMC is overkill and a dashboard tool tells you what you need.
AMC is Amazon's privacy-safe data clean room. It lets you run custom SQL queries against your own ad and retail data, including cross-channel overlap, path-to-purchase analysis, and audience builds that Amazon's native reports can't produce. Most brands use AMC to answer two questions: where is my ad spend actually driving incremental sales, and which audiences convert at the lowest cost. AMC is free if you run DSP; access without DSP requires a qualified seller-level partnership.
Holdout testing is the cleanest method: pause a specific campaign or ad type for 2–4 weeks on a controlled subset of products, hold all other variables constant, and measure the delta in total units sold against a matched control set. AMC's built-in incrementality reports automate the audience-holdout logic for DSP campaigns. Most brands that claim to "test incrementality" are actually measuring attribution shift. True incrementality requires a real holdout and patience to weather the short-term ranking dip.
Brand Analytics (BA) is a free Amazon Seller Central reporting suite with pre-built reports — search terms, market basket analysis, repeat purchase behavior — aggregated at the brand or search-query level. AMC (Amazon Marketing Cloud) is a privacy-safe data clean room where you run custom SQL queries against your own ad and retail data at the event level (subject to privacy aggregation thresholds). BA answers "what are shoppers searching for in my brand?"; AMC answers "across my DSP, Sponsored Ads, and retail events, what was the path to purchase for new-to-brand customers?" They're complementary. BA is good-enough for most Seller Central brands; AMC is what you graduate to when DSP spend or cross-channel attribution becomes a real question.
AMC applies a privacy-preserving aggregation threshold. You can't see results from any subset with fewer than ~100 unique users. Holdout tests need to design around this: either pick holdout markets large enough that both test and control groups clear the threshold, or structure your test as a pre/post analysis at the account level (which loses experimental clean-ness). For most mid-size brands, the cleanest approach is a geo holdout at DMA level. US is divided into 210 DMAs, and most DMAs have enough activity to clear the floor for DSP campaigns. Sponsored Products-only campaigns rarely have enough DMA-level granularity to run a clean holdout in AMC. Consider a market-matched attribution model outside AMC instead.
Limited but yes. Seller Central brands without DSP can access AMC through "AMC Lite", fewer queries, more aggregation, no first-party audience uploads. Full AMC requires DSP enrollment, which has $50K+ minimum monthly spend commitments. Most brands graduate to full AMC when their measurement questions outgrow Brand Analytics. Typically when they're investing $500K+/year in Amazon ads.
Custom audiences let you build specific shopper segments inside AMC and push them to DSP for targeting. Examples: "Shoppers who viewed your product but didn't buy in the last 30 days", retargeting. "Shoppers who bought your competitor's product 60+ days ago", conquest opportunity. "Shoppers who bought from your brand 12+ months ago". Win-back. Most brands use 5–10 audiences across the funnel; sophisticated programs run 30+.
Measure what ads really drive.
