Overview

Amazon Marketing Cloud (AMC) became generally available to advertisers in early 2023, offering sellers a secure, aggregated data platform for building and activating highly specific audience segments. Leveraging AMC enables advertisers to move past broad keyword or interest targeting, allowing them to allocate ad spend toward shoppers whose recent behavior aligns closely with the products they sell.

Key Points

• Privacy‑Centric Data — AMC processes only anonymized, aggregated shopper activity, so sellers never access personally identifiable information while still gaining actionable insights into purchase trends.
• Extended Event Window — Audiences can be defined using up to a 12‑month history of purchases, product views, and add‑to‑cart actions, such as “buyers who purchased premium coffee in the last 90 days and later viewed espresso machines.”
• Direct Campaign Export — Once an audience is generated, it can be pushed with a single click into Sponsored Display or Sponsored Brands campaigns, eliminating the need for manual list uploads.
• Lift‑Based Attribution — AMC supplies incremental sales and return‑on‑ad‑spend (ROAS) lift metrics that isolate the performance contribution of each custom audience.
• Iterative Testing Capability — Sellers can run A/B experiments on multiple audience definitions, compare conversion rates, and refine queries to pinpoint the most profitable shopper clusters.
• Scalable Audience Size — The platform reports estimated audience counts and confidence intervals, helping sellers gauge whether a segment is large enough to sustain a cost‑effective campaign.

How Amazon Marketing Cloud Works

1. Enter AMC via the Ads Console — After completing the Amazon Advertising enrollment, a seller logs into the Ads console and selects “Amazon Marketing Cloud.” For instance, an organic‑skincare brand clicks “Launch AMC” to unlock the query builder and data explorer.
2. Craft a Custom Audience with the Query Builder — Using a SQL‑style syntax, the seller writes a query that filters shoppers by event attributes. A typical example reads: This returns an aggregated group of pet‑product buyers who recently viewed a specific chew‑toy listing.