Audience Targeting and Segmentation
Building audiences
Amazon DSP targeting falls into a few families: in-market (shopping signals indicating active purchase intent), lifestyle (broader interest/behavioral signals), remarketing (advertiser-created from pixels, Advertiser Audiences, or Amazon shopper interactions), and third-party segments. Candidates should also know Amazon-created audiences versus advertiser-created audiences, and how Advertiser Audiences (hashed customer lists, pixel-based, and lookalikes) are built and refreshed.
How targeting combines
Line items let you layer in-market + lifestyle, apply exclusions (e.g., exclude recent converters), and use contextual targeting. Understand audience overlap and reach estimation - Amazon shows forecasted reach and may warn about narrow audiences. Frequency caps, recency (e.g., days since last visit), and lookback windows change who qualifies.
Traps
- Confusing in-market (purchase intent) with lifestyle (interest).
- Assuming remarketing requires only a pixel - Advertiser Audiences also accept hashed lists and lookalikes.
- Forgetting that exclusions reduce reach and that lookback windows affect segment membership.
- Treating Amazon shopper audiences as identical to third-party segments.
Worked example
A brand wants to reach people who viewed a product page in the last 30 days but did not purchase. Build a remarketing audience from the pixel with a 30-day lookback, then add an exclusion for purchasers (conversion pixel or Amazon shopper segment). Forecast reach will drop after exclusions - verify the line item still has enough scale before launch.