Amazon DSP — The Ultimate Extension of Your Amazon-PPC Ads — with Alexander Nelkenstock

There comes a terminal point in every successful Amazon brand’s lifecycle where Sponsored Ads simply run out of runway. You bid on the most relevant keywords, you optimise your bids, and you efficiently capture the existing demand. But what happens when you maximise the search volume in your category? You hit a ceiling. Brands that rely solely on standard pay-per-click models eventually find themselves trapped in a zero-sum game of rising bids, stagnant category growth, and fierce cannibalisation. For serious market players, merely harvesting existing intent is no longer a viable long-term strategy. Growth at scale requires creating new demand and relentlessly re-engaging shoppers who abandoned their carts to browse the wider internet.

This is the commercial void that Amazon DSP (Demand Side Platform) is built to fill. Far beyond the rigid confines of the Amazon search results page, DSP unlocks programmatic advertising across the open web, exclusive Amazon-owned properties, and premium video streaming environments. Yet, despite having been available for years, DSP remains profoundly misunderstood within the seller community. It is frequently dismissed as an over-priced branding luxury or, worse, mismanaged as a direct-response extension of Sponsored Display. Mastering the distinct mechanics, creative requirements, and unfamiliar metrics of this programmatic ecosystem is the ultimate dividing line between accelerating your brand and burning budget.

Why Sponsored Ads Alone Cap Your Brand

To understand the absolute necessity of DSP, one must categorise what standard Amazon advertising actually achieves. Sponsored Products and Sponsored Brands are fundamentally bottom-of-the-funnel tools. They rely on the cost-per-click (CPC) model and require a user to actively search for a keyword or navigate a specific category. They are designed for shoppers who are already holding their digital wallets.

However, standard PPC cannot proactively seek out a relevant demographic reading a news article or watching a stream. It cannot intelligently track a user who viewed your product, left Amazon, and is now reading a blog about home decor. Sponsored Display offers a rudimentary taste of remarketing, but its constraints are severe, and its attribution models can be notoriously chaotic. Without programmatic advertising, a brand is inherently passive, waiting for the customer to come to them.

Meet Alexander Nelkenstock and Xingu Advertising

To demystify this complex ecosystem, AMALYZE’s Christian Otto Kelm sat down with Alexander Nelkenstock from Xingu Advertising. Xingu operates as a specialist agency in the DACH region, focusing explicitly on full-funnel Amazon Ads and DSP execution. Navigating the DSP waters requires more than just a basic understanding of algorithmic bidding; it requires strategic architectural oversight.

The reality of Amazon DSP is that it functions as a highly granular, programmatic buying engine. As an agency that manages DSP for major brands, Xingu experiences first-hand the friction that occurs when brands approach programmatic buying with the restrictive mindset of an Amazon seller managing a PPC dashboard. The platform operates on an entirely different economic and technical foundation, necessitating a monumental shift in how advertising campaigns are planned, funded, and ultimately judged.

What Amazon DSP Actually Is (and Isn't)

At its core, Amazon DSP is Amazon’s programmatic display and video buying network. Instead of paying every time someone clicks on an ad, DSP is purchased on a CPM (cost per mille, or cost per thousand impressions) basis. You are buying guaranteed visibility and reaches, not guaranteed clicks.

The scope of this visibility is massive. A DSP campaign can serve ads on Amazon’s own premium properties, such as Twitch, IMDb, and Fire TV, but it also reaches across the open web via Amazon’s extensive publisher exchange. You might find a shopper looking at weather reports, scrolling through sports statistics, or reading a mainstream newspaper, and serve them a highly targeted Amazon advertisement based on their recent purchasing behaviour.

Crucially, DSP is not a blunt instrument. It is widely considered one of the most granular targeting platforms on the internet, powered by the most potent first-party retail data in the world. While Facebook knows what people 'like' and Google knows what people 'search', Amazon knows what people actually buy.

Three Audience Layers That Make DSP Work

The strategic architecture of a successful DSP campaign relies on layering audiences, typically structured from the bottom of the funnel upwards. The most immediate and lucrative layer is own-ASIN remarketing. This involves deploying ads to users who have interacted with your products in the past but failed to convert. You can target users who viewed your detail page, added an item to their basket, or, for consumable goods, purchased a product 30, 60, or 90 days ago and are now due for a replenishment.

The second layer is competitor and category targeting. This is where brands can act aggressively. By leveraging DSP, you can target shoppers who have viewed or purchased specific competing ASINs. For example, during the live discussion, a theoretical case study for an established brand like Edding acrylic markers was explored. Instead of competing head-to-head on generic search terms against inexpensive generic imports, Edding could target users who actively viewed low-cost alternatives, delivering a high-quality brand message to sway them. Alternatively, they could target users seeking companion products, such as crafting books or seasonal decorations.

The third and broadest layer involves Amazon's behavioural and lifestyle segments. Amazon places users into 'in-market' groups based on their collective browsing and purchasing history. This allows a brand to target macro-demographics—such as new parents, outdoor enthusiasts, or pet owners—with brand awareness campaigns to fill the top of the funnel before those users ever consider typing a specific product keyword into the Amazon search bar.

The Metrics That Confuse First-Timers

When brands transition from Sponsored Ads to DSP, the dashboard metrics often cause intense friction. The immediate instinct is to look for ACoS (Advertising Cost of Sales), but applying a direct performance metric to an upper-funnel display campaign is fundamentally flawed. DSP requires an entirely new vocabulary.

First, the attribution window is significantly longer. While Sponsored Ads typically default to a 7-day or 14-day click-based window, DSP heavily relies on a 14-day view-based window. Because it is an awareness and consideration tool, DSP claims attribution if a user simply views the ad and subsequently purchases the product within that timeframe, even if they never clicked the banner.

Second, brands must differentiate between Total ROAS and Promoted ASIN ROAS. DSP allows advertisers to see the 'halo effect' of their ads—sales of any product in the brand's catalogue. While a high Total ROAS looks impressive on a report, it often masks poor direct performance of the specific item being advertised. Instead, successful operators look at the Detail Page View Rate (DPVR) to see if the ad is actually driving traffic, and the New-To-Brand percentage (NTB%) to verify if the campaign is fulfilling its core mandate: acquiring fresh customers rather than merely cannibalising existing ones.

Creative: The Half of DSP Most Brands Ignore

A common misconception is that the targeting algorithm does all the heavy lifting. In reality, poor creative execution is responsible for a massive percentage of failed DSP engagements. Standard Sponsored Ads pull primary images directly from the listing. DSP, conversely, demands real advertising assets.

Brands need a robust pipeline of display banners, responsive eCommerce creatives (which dynamically feature product reviews, prime delivery badges, and current pricing), and increasingly, video assets. The introduction of Streaming TV (OTT) and non-skippable inventory on platforms like Twitch represents a seismic shift. Advertisers can now run 15- or 30-second broadcast-quality video ads in premium streaming environments. If a brand attempts to run DSP campaigns using generic white-background product shots cropped into standard internet banner sizes, the click-through rates will inherently suffer. Furthermore, the creative must explicitly signal that the click leads to Amazon, rather than an independent web shop, to meet user expectations and reduce bounce rates.

Minimum Budgets, Agency or Self-Service

The financial barrier to entry for Amazon DSP represents a significant hurdle for smaller sellers, functioning as a gatekeeper that ensures only maturing brands play in the programmatic space. Historically, self-service access directly through the Amazon Ads console required brands to commit to astronomical annual spending—frequently climbing well into the mid-six figures.

For the vast majority of brands aiming to enter the DSP space, the prevailing route is through an accredited agency. Agencies hold the primary seats and can partition access to individual brands. Even through this managed route, engaging effectively with DSP requires serious working capital. Deploying a budget of €200 for a month-long test is futile; the algorithm simply will not gather enough data to optimise impressions. An effective entry-level monthly budget generally begins around the €10,000 to €15,000 mark.

Regarding fee structures, standard agency models have traditionally sat at an additional 15 percent of media spend. However, as budgets increase, many expert agencies are shifting toward flat-tier retainers to ensure their fees correspond with the actual management labour rather than arbitrarily scaling with raw ad spend.

"A funnel is a funnel, and it always starts at the top and ends at the bottom. But 70 percent of consumers do not know what a funnel is, and 30 percent buy immediately by accident without waiting for consideration or awareness. You cannot micromanage the exact sequence of touchpoints; you have to run them in parallel and measure the holistic uplift."

When DSP Fails (and Why It's Rarely the Targeting)

Even with premium budgets and precise targeting, DSP campaigns can fail drastically. The root cause is rarely an algorithmic failure; it is almost always an alignment failure. The most common scenario occurs when a brand treats a DSP campaign exactly like a Sponsored Products campaign. They pour money into an upper-funnel display ad and then panic when the immediate ROAS appears negative after the first week.

Another frequent path to failure is underfunding a campaign while attempting to target a massive demographic. Spreading a modest budget across millions of potential impressions guarantees that the ad frequency remains far too low to drive behavioural change. Lastly, failing to isolate existing customers from acquisition targeting is a fatal flaw. Paying premium CPM rates to repeatedly serve lifestyle video ads to loyal customers who bought the product yesterday is the definition of budget incineration. Negative targeting—actively excluding recent purchasers—must be built into the foundation of the order line items.

Conclusion: DSP Is Brand Building With a Performance Backstop

Amazon DSP is not a magic bullet that will instantly double a brand's direct revenue without effort. It is a highly sophisticated, resource-intensive platform that demands strategic patience, excellent creative assets, and substantial capital. However, for a brand that has saturated its organic rank and maxed out its Sponsored Ads efficiency, DSP is the only legitimate vehicle for achieving the next stage of aggressive growth.

It bridges the gap between traditional brand building and data-driven performance marketing. It allows brands to tell a visual story across the open web, capture the attention of a distracted consumer on Twitch, and intercept a competitor's customer, all while anchoring those interactions to measurable retail outcomes over a 14-day window. If standard Amazon advertising is about catching a falling apple, DSP is about planting a new orchard.