If you add up Sponsored Products, Sponsored Brands and Sponsored Display separately and compare that against a combined total, the figures will not always reconcile cleanly. That is expected, and it is Amazon's doing rather than TrackIQ's.

[Watch the attribution walkthrough](https://kb.trackiq.com/kb-assets/kbvideo-attribution-windows-explainer-ca94a8d0.mp4)

*Prefer to watch? Fifty-one seconds on why the totals blend rather than sum.*

## **Different channels, different attribution windows**

Amazon credits a sale back to an ad for a fixed period after the click. That period is not the same for every channel:

| Channel | Attribution window |
|---|---|
| **Sponsored Products (SP)** | 7 days |
| **Sponsored Brands (SB)** | 14 days |
| **Sponsored Display (SD)** | 14 days |

So a combined, cross-channel view is a **blend of two different windows**. Sponsored Brands and Display get a fortnight to claim a sale; Sponsored Products gets a week.

![The headline tiles](/kb-assets/shot-ads-kpis-bee9d3.png)

## **What this means in practice**

**Do not compare SP efficiency directly against SB or SD efficiency.** Sponsored Brands has twice as long to be credited with a conversion. Its ACoS will tend to look better for that reason alone, before any real difference in performance.

**Recent days are unstable for SB and SD.** A Sponsored Brands campaign from three days ago has eleven days left to accumulate attributed sales. Its ACoS today will improve on its own. Judging a fortnight-window channel on three days of data will make it look far worse than it is.

**Totals shift as windows close.** A number you screenshotted last week can legitimately be different this week, because sales were still being attributed backwards into it.

## **The rule of thumb**

Compare like with like: SP against SP, SB against SB. When you do need one combined figure — for a monthly total, say — use a period that closed at least two weeks ago, so every channel's window has fully elapsed.

![The trend chart, where recent days are still filling in](/kb-assets/shot-ads-trends-bfc2fc.png)

## **One more thing that trips people up**

Campaign views default to showing **enabled** campaigns only. A month where you paused several campaigns will look like it lost spend that a full view would still show. If a total looks lower than you expect, check whether paused and archived campaigns are being counted.

Amazon DSP is a separate system again, with its own reporting. It is not part of the Sponsored Ads totals at all — see the **Amazon DSP** section for that.

## **Related**

- [**Understanding ROAS and Key Amazon Advertising Metrics**](https://kb.trackiq.com/hc/knowledge-base/articles/understanding-roas-and-key-amazon-advertising-metrics) — the metrics themselves.
- [**Time to Conversion: Checking Your Attribution Window**](https://kb.trackiq.com/hc/knowledge-base/articles/time-to-conversion) — the attribution windows behind the overlap.
- [**Spot Cannibalisation Between Sponsored Brands and Sponsored Products**](https://kb.trackiq.com/hc/knowledge-base/articles/spot-sb-sp-cannibalisation) — the playbook for testing real overlap.
- [**Create a Sponsored Brands Video Campaign**](https://kb.trackiq.com/hc/knowledge-base/articles/create-a-sponsored-brands-video-campaign) — the fourth thing people try to add in.
- [**Ad Type and Placement Mix**](https://kb.trackiq.com/hc/knowledge-base/articles/ad-type-and-placement-mix) — rebalancing the mix without adding the ad types together.