Introduction
Amazon Ad Server (formerly Sizmek) is a third-party ad-serving platform used to deliver, measure, and verify display, video, and rich-media creatives across the open web. The connector pulls delivery, engagement, video, cost, and conversion data across the ad account, campaign, ad set, and ad hierarchy. As a marketing connector, metrics, breakdowns, and currencies are standardized so widgets can aggregate across sources without additional configuration. To connect this data source, see How to connect Amazon Ad Server to Adriel.Data refresh strategy
Architecture data
Architecture covers ad accounts, campaigns, ad sets, and ads. It refreshes twice daily at 4:00 PM UTC and 5:00 AM UTC.Reports data
Reports include daily performance for ad accounts, campaigns, ad sets, and ads. Cache-only. Report data comes from cached daily performance snapshots. When a dashboard loads, data is composed from the last snapshot rather than fetched on demand, so uniqueness-based metrics such as reach may be overvalued across date ranges longer than one day. Initial fill. On data source creation, the cache fills with 365 days of historical data at the ad account, campaign, ad set, and ad levels. Refresh schedule. Reports refresh on two overlapping schedules:- 9:00 PM and 9:00 AM UTC daily — syncs the last 9 days for ad account, campaign, ad set, and ad
- 6:00 PM UTC every 2 days — syncs the last 17 days for ad account, campaign, ad set, and ad
Architecture levels
Ad account → Campaign → Ad set → AdDate range limits
Metrics
How to read the API Key columnCode style like
Impressions (Net) is the literal Amazon Ad Server metric name requested from the platform. Italic text describes how a value is produced when it doesn’t map cleanly to a single platform field (Adriel-computed rates or values aggregated internally).Performance
Video performance
Conversion performance
Conversions are fetched in a separate report and merged into row data. Each conversion event surfaces its own count and value, along with derived CPA, ROAS, and CVR. Metric names below follow the pattern for each event, where[Conversion event] is the event’s name.
Breakdowns
Standard hierarchy
Creative & placement
Geographic
Overview & status
Time & date grouping
The platform returns daily records; Adriel groups them into these views for reporting. All time-grouping breakdowns are Adriel-side aggregations of the same underlying daily data.
Limitations
Based on the current connector, Amazon Ad Server has the following limits:- Cache-only data — all report data is served from scheduled daily snapshots. Uniqueness-based metrics such as reach may be overvalued when viewed across date ranges longer than one day, and data outside the cache sync window may be incomplete.
- Unique metrics need a breakdown — Unique Clicks and Unique Impressions require at least one of Ad, Campaign, or Placement as a breakdown. They are unavailable at the ad-account level, where none of these are present.
- Total Custom Interactions conflict — Total Custom Interactions cannot be requested at the same time as conversions, so it is excluded from the default request set.
- Placement breakdown availability — the Placement (Site) breakdown is limited to specific workspaces and is not enabled by default.
- Two parallel reports — every fetch fires a basic performance report and a conversions report in parallel, merged before normalization. This makes the connector slower to refresh than single-report integrations, and a partial result may appear if one report fails while the other succeeds.
- Architecture volume cap — ad discovery reads pages of 500 ads and stops with an error if it exceeds roughly 50,000 ads without completing.
API references
- Amazon Ad Server SAS API (
adapi.sizmek.com/sas/) — architecture data and authentication - Amazon Ad Server Report Builder API (
api.sizmek.com/rest/ReportBuilder/reports/) — asynchronous report generation
See also
- How to connect Amazon Ad Server to Adriel — step-by-step setup guide
- Amazon DSP data reference
- Amazon Ads data reference
