Key Takeaways
- Amazon’s default Seller Central reporting uses last-click attribution, which credits the final ad a customer clicked before purchasing and ignores every prior touchpoint.
- DSP campaigns consistently show understated ROAS in standard reporting because they operate earlier in the buyer journey, where last-click attribution assigns them no credit.
- Amazon Marketing Cloud (AMC) tracks the full path to purchase using first-party data from over 300 million active customers, revealing which ad combinations actually drive conversions.
- In a real 7-figure account, DSP reported 12 sales on $4,000 in spend; AMC revealed the same campaigns had influenced 47 conversions credited elsewhere, producing an actual ROAS of 3.2x.
- Buyers who saw DSP first and then Sponsored Products convert 19% higher than buyers who only saw Sponsored Products.
- A 3-phase implementation framework covering foundation, controlled testing, and scale brings DSP to a 2x blended ROAS within 8 to 10 weeks for most established brands.
General Summary
Amazon DSP campaigns routinely look like failures in Seller Central because the platform’s default reporting credits only the last ad a customer clicked before purchasing. Every awareness impression, every consideration touchpoint, every display view that built intent before the final click disappears from the record. The result is a systematic distortion: Sponsored Products campaigns appear to carry the full weight of conversions they did not start, while DSP campaigns that initiated the customer journey receive zero credit. Amazon Marketing Cloud resolves this by tracking anonymized users across every touchpoint in sequence, using Amazon’s own first-party data. When brands run the Path to Purchase report in AMC, they typically find that DSP influenced 35 to 45% of conversions attributed to other campaigns in Seller Central. The practical consequence is that brands spending $30,000 to $100,000 per month on DSP are frequently cutting or pausing campaigns that are actively feeding their best-performing Sponsored Products. A 3-phase framework using AMC measurement, controlled budget allocation, and audience refinement brings DSP performance into view and creates the conditions for profitable scaling.
Extractive Summary
Amazon’s default attribution credits the last ad clicked before a purchase, making DSP campaigns appear to underperform relative to Sponsored Products. DSP operates primarily in awareness and consideration stages, where buyers view ads without clicking before later searching and buying through a different ad unit. Amazon Marketing Cloud tracks the full path to purchase using anonymized first-party data from Amazon’s customer base, connecting DSP impressions to downstream conversions that standard reporting never links. A home goods brand spending $4,000 on DSP reported 12 attributed sales in Seller Central; AMC data showed those campaigns had influenced 47 conversions credited to Sponsored Products, reflecting a true ROAS of 3.2x. Buyers exposed to DSP before Sponsored Products convert 19% higher than buyers who only see Sponsored Products. A 3-phase implementation covering AMC setup, a $10,000 to $15,000 controlled test, and audience-level scaling produces measurable results within 60 to 90 days for brands with stable Sponsored Products performance.
Abstractive Summary
The DSP attribution problem is a measurement problem masquerading as a performance problem. When brands cut DSP because it shows low ROAS, they are not responding to evidence of failure; they are responding to a reporting gap. The gap exists because last-click attribution was designed for a single-channel world, where the last ad seen was often the only ad seen. In a multi-channel environment with display, video, streaming, and search all running simultaneously, last-click attribution systematically favors the channel closest to the conversion moment, which is almost always Sponsored Products. DSP’s role is fundamentally different: it creates the conditions for Sponsored Products to convert by building recognition and intent before the search happens. Cutting DSP to protect Sponsored Products efficiency is the equivalent of cutting your sales team’s prospecting budget because the closers have higher conversion rates. The closers’ numbers look good precisely because someone did the earlier work. AMC does not change how DSP performs; it changes what brands can see. And what they see, consistently, is that the campaigns they were about to kill were doing more work than any other line in the account.
Why Does Amazon DSP Look Like It’s Not Working?
Amazon DSP appears to underperform because Seller Central uses last-click attribution, which assigns 100% of the conversion credit to the final ad a customer clicked before purchasing. Every earlier touchpoint, including the DSP impression that first created awareness of the product, receives no credit at all.
The pattern is consistent across accounts. A seller opens Amazon Ads and checks their display campaigns. CPM sits around $15. Click-through rate is acceptable. But ROAS shows 1.2x or 1.5x while Sponsored Products campaigns sit at 4x or 5x. The conclusion looks obvious: shift budget away from DSP.
The conclusion is wrong. DSP campaigns operate at the top and middle of the funnel, where buyers build awareness and consideration before they ever search. A buyer who sees a DSP video ad on Prime Video does not click immediately. They continue watching their show. Two days later, they search the product category, see a Sponsored Products ad, click, and buy. Seller Central credits Sponsored Products for the full conversion. The DSP ad that started the journey gets nothing.
The more expensive the product, the more severe this distortion becomes. Higher-priced items require longer consideration periods, more touchpoints, and more research before purchase. Each additional touchpoint is another opportunity for DSP to influence the outcome invisibly.
What Does Last-Click Attribution Actually Hide?
Last-click attribution hides the contribution of every ad in the buyer journey except the final one, which means it systematically undervalues awareness and consideration campaigns while overstating the independent contribution of conversion-stage campaigns.
In a typical multi-touchpoint journey, a buyer might see a DSP display ad on a third-party website, visit the brand store through a Sponsored Brands ad three days later, and then search and click a Sponsored Products ad on day five before purchasing. Seller Central records one conversion for Sponsored Products. DSP and Sponsored Brands show zero conversions for this buyer, despite each playing a measurable role.
The consequence for budget allocation is direct. Brands reading Seller Central data move money away from the channels that create demand and toward the channel that harvests it. Sponsored Products conversions hold steady for a few weeks. Then they drop. The awareness pipeline has dried up.
One brand cut $30,000 in DSP spend after seeing poor ROAS in Seller Central. Three weeks later, their Sponsored Products conversions fell 35%. The DSP campaigns had been feeding the SP pipeline the entire time. The connection was invisible without the right measurement tools.
What Is Amazon Marketing Cloud and How Does It Work?
Amazon Marketing Cloud (AMC) is a cloud-based measurement environment that tracks anonymized customer interactions across every Amazon ad touchpoint in sequence, using Amazon’s first-party shopping data rather than cookies or third-party estimates.
AMC connects to the actual purchase behavior of over 300 million active Amazon customers. When a buyer interacts with a DSP impression on Monday, visits a brand store on Wednesday, clicks a Sponsored Products ad on Friday, and purchases on Saturday, AMC records all four events as a single connected path.
Seller Central shows who scored the goal. AMC shows every pass that led to it.
AMC operates through SQL-based queries run against anonymized event-level data. Amazon provides a pre-built Instructional Query Library with templates for common analyses. The Path to Purchase template requires about 45 seconds to set up and returns data on which ad sequences preceded conversions, how long the journey took, and which touchpoints appeared most frequently in converting paths.
What Does AMC Reveal That Seller Central Misses?
AMC consistently reveals that DSP campaigns influence 35 to 45% of conversions that Seller Central credits entirely to other campaigns, particularly Sponsored Products.
The Path to Purchase report shows the exact sequence of ad exposures that preceded each conversion. For most accounts, DSP appears early in the majority of converting paths, establishing awareness and intent that downstream campaigns then convert. Without AMC, this contribution is completely absent from any reporting a brand sees.
AMC also surfaces the sequence effect. Buyers who saw a DSP ad first and then a Sponsored Products ad convert 19% higher than buyers who only saw the Sponsored Products ad. That 19% lift represents the measurable value DSP adds to every Sponsored Products conversion it precedes. Last-click attribution assigns none of that value to DSP.
A second finding AMC surfaces regularly is frequency waste. Brands often find that 30 to 40% of their DSP budget is going to users who have already seen their ads 15 or more times without converting. Capping frequency at 5 to 7 impressions per week and redirecting that budget to fresh audiences produces ROAS improvements of 20 to 25% without adding spend.
What Did AMC Find in a Real 7-Figure Account?
In a home goods brand generating $2 million per month on Amazon, DSP display campaigns showed $4,000 in spend and 12 attributed sales in Seller Central, a reported ROAS of 0.8x. The account team was preparing to cut the campaigns.
The AMC Path to Purchase report showed a different picture. Those 12 reported sales were the conversions where a buyer happened to click a DSP ad last. The same campaigns had appeared in the path of 47 additional conversions that Seller Central credited to Sponsored Products.
Adding those 47 conversions to the 12 already attributed raised the true conversion count to 59. Against $4,000 in spend, the actual ROAS was closer to 3.2x. The campaign the team was about to cut was one of the strongest performers in the account.
The same report identified the frequency problem. 40% of DSP budget was going to users who had seen the ads more than 15 times. After capping frequency at 5 to 7 impressions per week and reallocating the freed budget to new audiences, ROAS improved by a further 25%.
The campaigns that looked like failures were driving 40% of the account’s conversions. The campaigns that looked like winners were, in significant part, harvesting demand that DSP had created.
How Do You Set Up AMC to Measure DSP Accurately?
Setting up AMC for accurate DSP measurement starts with access: brands spending on Amazon Ads can request AMC access directly from Amazon, or through their agency if one manages the account.
Once in AMC, navigate to the Instructional Query Library and run the Path to Purchase template. The template is pre-built and takes roughly 45 seconds to configure. The output shows which ad types appeared in converting paths, in what sequence, and how frequently DSP appeared without receiving last-click credit.
The key number to identify first is the percentage of total conversions where DSP appeared in the path but did not receive last-click attribution. In most accounts audited, this sits between 35 and 45%. That percentage is the measurement gap: the share of DSP’s actual contribution that Seller Central cannot see.
After the path report, run a frequency analysis to identify users receiving 10 or more impressions within a 7-day window. This segment typically consumes budget without contributing conversions and represents the most immediate reallocation opportunity.
What Is the 3-Phase Framework for Running DSP Profitably?
A 3-phase framework covering foundation, controlled testing, and scale brings DSP to a 2x blended ROAS within 8 to 10 weeks for brands with stable Sponsored Products performance. The framework takes 60 to 90 days total.
Phase 1: What Does the Foundation Stage Require?
The foundation stage (Weeks 1 to 4) establishes measurement infrastructure before allocating DSP budget. Without AMC configured and baseline data collected, any DSP spend produces results you cannot accurately evaluate.
Request AMC access, run the Path to Purchase query, and document your current attribution picture. Identify what share of your existing conversions already have DSP touchpoints in the path. This baseline becomes the comparison point when DSP campaigns launch or scale.
Confirm that Sponsored Products are stable before adding DSP. If SP ACoS exceeds 35% or the account is generating fewer than 100 conversions per month, fix the SP foundation first. DSP amplifies what is already working; it does not compensate for weak conversion infrastructure.
Phase 2: How Should the Controlled Test Be Structured?
The controlled test phase (Weeks 5 to 12) runs $10,000 to $15,000 in total DSP spend across 60 to 90 days, split across 3 funnel stages with different KPIs for each.
Allocate 30% to awareness: video ads and streaming placements on Prime Video and Twitch. The KPI for awareness is reach and video completion rate, not ROAS. Judging awareness spend on immediate purchase returns is the same error that makes DSP look bad in Seller Central.
Allocate 35% to consideration: retargeting product page viewers and brand store visitors. These buyers have already shown intent. The KPI is click-through rate and brand store visit rate.
Allocate 35% to conversion: cart abandoners and high-intent shoppers who have viewed the product 3 or more times. This stage can be evaluated on ROAS because the buyer is already close to purchase. Conversion-stage DSP typically shows the strongest last-click numbers and serves as the anchor that makes the overall campaign look defensible in Seller Central even before AMC is used.
Phase 3: When and How Do You Scale DSP?
The scale phase (Week 12 onward) begins when AMC data confirms a 2x blended ROAS across all 3 funnel stages combined. Most brands reach this threshold by weeks 8 to 10 if the foundation and test phases are executed correctly.
At scale, target 15 to 20% of total Amazon ad spend allocated to DSP. Build custom AMC audiences based on specific behavioral signals: cart abandoners who did not purchase within 7 days, buyers who viewed the product 4 or more times without converting, and frequency-capped segments that exclude users approaching ad fatigue.
Continue running weekly AMC queries to monitor path composition, frequency distribution, and cross-campaign influence. The measurement infrastructure built in Phase 1 is not a setup step; it is the ongoing operating system that keeps DSP allocation calibrated as the account grows.
What Should You Check Before Making Any DSP Budget Decision?
Before cutting, pausing, or scaling any DSP campaign, run the AMC Path to Purchase report and identify what percentage of total conversions had DSP in the path without receiving last-click credit. That number is the minimum adjustment to make before any budget decision.
If that number is above 30%, the Seller Central ROAS figure for DSP is materially misleading. The actual contribution is higher. Cutting the campaign on the basis of the Seller Central figure means cutting a campaign that is producing results you cannot see.
Check frequency distribution next. If a significant share of impressions is going to users who have already seen the ad 10 or more times, reallocation to fresh audiences will improve performance without increasing spend. This is the fastest lever available once AMC is configured.
The brands that run DSP profitably at scale are not necessarily spending more or running more sophisticated creative. They are measuring more accurately. AMC closes the gap between what is happening in the account and what the standard dashboard shows. That gap, in most 7-figure accounts, is where the real performance story lives.
