Amazon DSP Category-Level Omnichannel Metrics: A Performance Marketer Playbook
What launched on April 14
Amazon rolled out product category breakouts inside DSP Omnichannel Metrics for the US market on April 14, 2026. The reports now split DSP-attributed sales by ASIN category taxonomy, which is the first time a DSP buyer can tie upper-funnel display or streaming spend back to specific SKU clusters without stitching data manually.
Before this, DSP sales sat in aggregate or campaign-level buckets. You knew the campaign drove 400 units at a 3.1 ROAS, but not whether those units were baby bottles, cookware, or patio chairs. For single-line brands that gap rarely mattered. For anyone running a mixed catalog, it forced budget calls on blended averages that hid a lot.
Nine days later, on April 23, Amazon Ads announced an expansion of the Acxiom audience integration in DSP Audience Hub to the UK and Germany. That brings the same 10,000-plus Acxiom audience segments US buyers already had into the two biggest EU retail media markets. The category report and the audience library land in the same quarter, which is not a coincidence.
Why aggregate DSP attribution was painful
The previous reporting was accurate, just too coarse. If DSP served a display impression and the shopper later bought anything in your catalog within 14 days, that sale counted. Fine in theory. In practice a star category absorbs the credit and drags the average up while a weaker line burns quietly.
The attribution window compounds the issue. DSP Display defaults to 14-day view-through plus 1-day click. Streaming TV goes out to 28-day view-through. Wider windows make blended metrics look great while obscuring which categories DSP is actually moving versus which ones would have converted from organic search anyway.
Category breakouts do not change the windows. They change the slice. You still get the same counted sales, now distributed across the categories those sales occurred in, which is enough to surface the structural imbalances that aggregate reporting hides.
Four concrete use cases
Week one with the new report should not be about budget changes. It should be about diagnosis. Run through this grid before touching anything.
| Use case | Action | What you get |
|---|---|---|
| Diagnose budget misallocation | Pull DSP spend by category next to DSP attributed sales by category | Categories that are over or underfunded relative to the revenue they produce |
| Prune low-performing spend | Pause DSP for categories with ROAS more than 30 percent below account average for 2 weeks | Confirm the category is actually dragging the account rather than running a seasonal lull |
| Compare DSP vs Sponsored Products | Lay DSP attributed sales next to Sponsored Products sales within the same category | Decide if that category needs upper-funnel push or if search traffic already covers it |
| Build high-AOV retention audiences | Export high-AOV buyers from a specific category via AMC, feed them to DSP as a seed | Stronger lookalikes and retargeting pools for next quarter’s launches or anniversary pushes |
A worked example
A US merchant I worked with runs three product lines through DSP: kitchen, outdoor, and baby. Budget had been split evenly across the three, roughly 33/33/33, for the prior six months. Aggregate ROAS was sitting at 3.2. Nothing in the old reports suggested any category was off, because the old reports could not show it.
Week two of the new breakouts changed the picture. Baby SKUs were driving 70 percent of DSP-attributed revenue. Kitchen pulled 22 percent. Outdoor produced 8 percent despite taking a third of the budget. A third of DSP spend was going into a line that converted poorly on upper-funnel inventory.
We rebalanced to 60/20/20 across baby, kitchen, and outdoor. Outdoor did not get cut to 5 percent because the category was entering its natural season and we did not want to starve it just before demand picked up. Thirty days in, blended ROAS moved from 3.2 to 3.9, a 22 percent lift. Outdoor impressions dropped significantly but conversions barely moved, which confirmed the old allocation was mostly wasted reach rather than incremental sales.
UK and Germany teams: what the Acxiom expansion means
If you are running DSP out of London or Munich, the Audience Hub update is the bigger of the two launches. Acxiom segments cover demographic, life-stage, and financial behavior cuts that Amazon’s first-party audiences do not capture. Pairing those with the new category breakouts gives EU teams a toolset the US side has had for about a year.
Two practical notes. Audiences do not auto-activate into existing line items. Someone has to go into Audience Hub, pick the segments, and attach them. Agency handoffs in particular should not assume the previous manager did this. Second, do not copy the US activation list wholesale. UK grocery habits, German home appliance shopping patterns, and household composition cuts all differ, and Acxiom’s local segmentation reflects that.
Caveats before you start reallocating
Attribution windows have not changed. Display is still 14-day view-through plus 1-day click, Streaming TV stretches to 28-day view-through. If you had philosophical issues with view-through credit before, the new report inherits them. Category assignment also uses the ASIN’s Amazon browse node, not your seller-defined storefront categories, so mislabeled ASINs produce weird category totals. If a line looks surprisingly small or large in the report, check the browse node mapping before reading anything into the number.
AMC audience builds need about 14 days of data behind them. A brand new DSP account will see real-time category sales immediately but cannot feed meaningful lookalike seeds to DSP until enough behavior has accumulated. Plan the new-launch calendar around that lag instead of against it.
One more note, judgment rather than mechanics. Category breakouts reveal more decisions, but not every decision needs to happen the week the report lands. Let 30 days of baseline accumulate before making structural cuts, especially in categories with strong seasonality.
阅读本文中文版: 亚马逊 DSP 全渠道指标拆到品类：卖家怎么用这份新报表重新分配预算
Related Articles
Google Ads Promotion Mode: Peak-Season Bidding for Search and Performance Max Without the Rollback Risk
In mid-June 2026 Google Ads launched Promotion Mode (beta) for Search campaigns (including AI Max features) and Performance Max, letting advertisers temporarily loosen ROAS tolerance and add budget during a promo window with automatic rollback at the end. When cross-border sellers should use it, when to skip it, and how it differs from manual budget and bid changes.
TikTok Asset Manager + Auto Selection: How Creative Workflows Actually Change
TikTok's Asset Manager and Auto Selection, announced at TikTok World 2026, target a different problem than GMV Max Pro or Search Hubs: they centralize catalog, data connections, and creative in one hub, then auto-pick the best-predicted assets across creator content, product assets, and Symphony-generated creative. Here is what changes in a seller's day-to-day creative workflow.