Amazon DSP Measurement & Attribution
The platform goes beyond ROAS by offering sophisticated metrics like New-to-Brand (NTB) and Brand Halo to demonstrate true business growth, along with tools like Amazon Marketing Cloud (AMC) for in-depth data analysis.
I. The Fundamentals of Amazon Attribution
Unlike other platforms, Amazon DSP uses "closed-loop" measurement because they own both the ad tech and the retail platform.
- Attribution Type: 14-day, last-touch attribution.
- The Hierarchy of Credit: 1. Click-Through (Highest Priority): If a user clicks an ad and buys within 14 days. 2. View-Through: If a user views an ad (at least 50% on screen for 1 second) and buys within 14 days without clicking.
- The "Click Over View" Rule: If a user sees a Video ad (view) but then clicks a Display ad, the Display ad gets 100% of the credit.
II. Sophisticated Measurement Capabilities
The guide highlights several advanced ways Revive Media tracks performance beyond simple ROAS:
1. New-to-Brand (NTB) Metrics
- Definition: Purchases made by customers who haven’t bought from your brand on Amazon in the last 12 months.
- Value: This is the primary KPI for "Upper Funnel" campaigns. It distinguishes between defending your brand and growing it.
2. Retail Insights (The "Halo" Effect)
- Promoted vs. Brand Halo: * Promoted: Sales of the exact ASINs in the creative.
- Brand Halo: Sales of other products in your catalog.
- Significance: This demonstrates how DSP ads for a "hero product" often lift the sales of the entire brand portfolio.
3. Customer Long-Term Value (CLV)
- The DSP allows for tracking "Subscribe & Save" sign-ups triggered by ads, which helps calculate the long-term ROI of an initial ad-driven purchase.
III. Amazon Marketing Cloud (AMC) Integration
This is a critical section of the Revive Media guide. AMC provides a "Clean Room" for deeper data:
- Cross-Channel Overlap: See how many customers interacted with both a Sponsored Product (Search) and a DSP ad (Display) before buying.
- Path to Purchase: Analyze the average number of impressions it takes for a customer to convert.
- Custom Attribution: Ability to build "Multi-Touch Attribution" (MTA) models that give credit to every touchpoint, not just the last one.
IV. Offline & Third-Party Measurement
For brands that sell outside of Amazon or have specific brand-lift goals:
- Amazon Brand Lift Studies: Surveying exposed vs. unexposed audiences to measure "Top of Mind Awareness" and "Purchase Intent."
- Amazon Lead Generation: Measuring sign-ups or form fills directly within a DSP ad unit.
- Omnichannel Impact: Using Amazon's partnership with companies like Whole Foods to see if DSP ads drove in-store grocery purchases.
V. Strategic Implementation Checklist
To maximize these capabilities, the guide suggests following this framework:
- Define the Window: While 14 days is standard, analyze the "Time to Convert" report to see if your customers actually take longer.
- Set Tiered KPIs: * Awareness: Reach, Frequency, Viewability.
- Consideration: DPV (Detail Page View) Rate, ATCR (Add to Cart Rate).
- Conversion: ROAS, NTB, Unit Sales.