Adelaide today launched attention-based pre-bid targeting inside Amazon DSP, giving programmatic buyers a way to apply its AU media quality metric at the point of activation - before any spend is committed.

Adelaide, the attention measurement company, has made its AU pre-bid segments available in Amazon DSP, extending the reach of its outcome-based media quality metric into one of the advertising industry's most widely used demand-side platforms. The launch marks a meaningful shift in how attention data can be applied: moving it from post-campaign reporting into active bid decisioning.

What AU pre-bid targeting does inside Amazon DSP

Until now, advertisers using Adelaide alongside Amazon DSP could access impression-level visibility into attention quality, but only after a campaign had already run. The new pre-bid capability reverses that sequence. According to Adelaide, buyers can now prioritize or exclude inventory based on AU scoring before they bid, rather than reviewing results after the budget has already been spent.

The segments are built from historical placement-level AU data. Adelaide groups supply according to its likelihood of capturing attention and driving downstream outcomes, drawing on cumulative, statistically significant observations across placements. The methodology is designed to reflect repeatable patterns in media quality rather than isolated spikes in performance - a distinction that matters when using historical data to make forward-looking buying decisions.

Inventory coverage across the integration is broad. Buyers can apply the segments across Display, Online Video, and Streaming TV inventory within Amazon DSP, spanning desktop, mobile web, and app environments. The combination of format types and device environments makes the integration applicable to a wide range of campaign structures.

The two products launching in Amazon DSP

The launch packages Adelaide's attention intelligence into two distinct products, each targeting a different buying objective.

AU Media Quality segments inventory into three performance tiers - High, Average, and Low - based on AU scores. Buyers can target the High tier to prioritize placements most likely to capture attention and influence outcomes. The Average tier is positioned as a balance between quality and scale, useful for campaigns where reach constraints require access to a broader pool of inventory. The Low tier serves as an exclusion layer, allowing buyers to actively filter out placements at the bottom of the attention quality range. According to Adelaide, segments can also be combined to accommodate different goals and scale requirements within a single campaign.

AU Quality Floor is designed as a simpler, single-step waste exclusion filter. It allows buyers to exclude both MFA-designated (made-for-advertising) sites and the bottom 10% of AU-scored placements in one action. The segment launches exclusively inside Amazon DSP. According to Adelaide, it is intended for advertisers who want to establish a quality minimum without configuring more granular tier-based rules.

The MFA exclusion element is notable given the scale of the problem. DoubleVerify's 2024 Global Insights Report documented a 19% year-over-year increase in MFA impression volume in 2023, analyzing more than one trillion impressions from over 2,000 brands globally. The ANA found that 15% of programmatic spend flows to MFA websites. Combining that classification with AU scoring in a single segment provides a consolidated quality baseline that would otherwise require multiple separate filters.

What Adelaide and Amazon said

Marc Guldimann, CEO and Co-founder of Adelaide, framed the launch around timing - specifically, the problem of learning about media quality only after a campaign is over.

"Advertisers shouldn't have to wait until after a campaign runs to know whether their spend went toward high-attention media," said Guldimann. "By making AU Media Quality available within Amazon DSP, we're giving buyers a more practical way to apply attention-based standards during activation, before spend is committed."

Chris Conetta, Director of Omnichannel Supply at Amazon DSP, addressed the practical demand from buyers for more control at the activation stage.

"Advertisers are looking for more ways to increase efficiency and reduce media waste when purchasing media," said Conetta. "The availability in Amazon DSP for buyers to have additional control at the point of activation helps them align bidding decisions with their media quality standards."

Neither statement quantifies performance improvements buyers can expect from applying the segments, and no test data accompanies the launch announcement.

How AU works as a metric

The AU metric - Adelaide's core product - is an omnichannel measure of a placement's likelihood of capturing attention and driving outcomes. It does not rely on a single signal. According to Adelaide, it analyzes factors including ad size, time in view, clutter, and position to model the probability that a given placement will register with an audience and generate business results. The metric is designed to predict full-funnel outcomes, from awareness through conversion, rather than measuring a single point in the customer journey.

Adelaide has previously described the metric as outcome-based rather than purely exposure-based. Standard viewability, by contrast, only confirms that an ad met a technical threshold for on-screen visibility - 50% of pixels visible for at least one second for display, two seconds for video - without measuring whether a user actually engaged with what they saw.

The pre-bid application builds on that distinction. A viewability filter, such as the Oracle Data Cloud Pre-Bid by Moat integration Amazon offered previously, sets a binary threshold - the placement was or was not viewable above a given rate. AU pre-bid segments layer quality gradations on top of that, enabling buyers to make decisions based on predicted attention performance rather than minimum technical exposure.

Amazon DSP previously integrated DoubleVerify, Integral Ad Science, and Oracle Data Cloud for pre-bid targeting around invalid traffic, viewability, and brand safety. Adelaide's integration extends the category of pre-bid signal types available on the platform to include outcome-based attention quality.

Context: Adelaide's expanding programmatic footprint

The Amazon DSP launch is the latest in a series of partnerships Adelaide has built across programmatic infrastructure. The pattern is consistent: Adelaide positions its AU metric as a tool that works within existing buying workflows rather than as a standalone platform requiring separate access.

In June 2025, Comscore expanded its Certified Deal IDs to include Adelaide's attention metrics through PubMatic, combining Comscore's independent inventory quality signals with AU scoring across PubMatic's premium supply-side inventory, including connected TV publishers. That deal created curated inventory packages where buyers could access supply certified by both Comscore's content quality rankings and Adelaide's attention performance criteria.

In October 2025, Nielsen and Adelaide announced an integration combining AU with Nielsen's reach data inside the Outcomes Marketplace within Nielsen ONE. The collaboration gave advertisers simultaneous visibility into how many people saw an ad and how effectively that media captured attention - two dimensions that had previously required separate reporting from separate vendors.

In October 2025, Uber Advertising partnered with Adelaide and Kantar to develop a Custom AU metric calibrated to Uber's advertising environment. That collaboration involved platform-specific attention measurement developed through structured brand outcome research.

These partnerships indicate Adelaide is pursuing a strategy of embedding AU into the infrastructure buyers already use, rather than asking advertisers to adopt new tools. The Amazon DSP integration fits that model directly.

Why pre-bid matters more than post-campaign reporting

The shift from post-campaign to pre-bid is not a minor product update - it changes what the data can actually do for a buyer. Post-campaign AU reporting tells advertisers how their spend distributed across quality tiers after the fact. It supports future planning but does not change what happened in the campaign being measured. Pre-bid segments let buyers set quality floors and ceilings before the auction, so bidding decisions reflect media quality criteria from the start.

This has particular relevance in environments with large available inventory pools. Amazon DSP reached a monthly ad-supported audience of 275 million customers in the United States as of October 2024, and the platform has expanded across Netflix, Spotify, Disney, and other premium publishers. At that scale, the manual curation of placements becomes impractical. Automated pre-bid filtering based on AU scores offers a way to apply quality criteria systematically, without requiring campaign managers to individually assess inventory sources.

The AU Quality Floor product addresses the most basic version of this problem: removing placements that are unlikely to contribute to outcomes without requiring buyers to engage with tier configuration at all. The description of it as a "single-click quality minimum" reflects an intent to lower the barrier to quality filtering for buyers who may not have the resources to manage more granular approaches.

The MFA component of AU Quality Floor addresses an ongoing structural issue. MFA sites account for approximately 15% of programmatic ad spend, according to ANA data, and they often pass basic viewability checks while delivering poor brand lift and conversion performance. Combining MFA exclusion with the bottom-decile AU filter means the segment targets two distinct but overlapping sources of low-quality inventory.

What this means for media buyers and the attention market

Attention measurement has been moving toward activation for some time. The July 2025 IAB Europe Virtual Programmatic Day panel found attention metrics increasingly influencing programmatic decisions, but noted that measurement standardization remained a challenge across providers. Pre-bid integration does not solve the standardization problem - AU remains a proprietary metric rather than an industrywide standard - but it does make the metric actionable at a point in the buying process where it can affect budget allocation directly.

For CTV buyers specifically, the streaming TV component of the Amazon DSP integration is relevant context. The CTV measurement landscape has been developing rapidly, with deterministic outcome measurement, show-level transparency, and attention scoring tools all arriving within roughly the same 12-month window. Over-one-third of CTV ad impressions are delivered in TV-off environments, contributing to an estimated $1 billion in annual wasted advertising spend, according to industry figures. Applying attention-based pre-bid filters to Streaming TV inventory within Amazon DSP adds another quality layer to a format category where waste has been a documented problem.

The programmatic context matters too. Amazon's auto deal selection feature, launched April 15, 2026, uses machine learning to build curated inventory groups for Streaming TV campaigns based on campaign objectives. That automation addresses curation at the deal level. AU pre-bid segments operate at the placement level, applying quality scoring to individual impressions rather than pre-packaged deal pools. The two approaches can coexist within the same campaign structure.

Programmatic advertising growth reached 72% according to industry data, and the platforms buyers use to access that inventory are accumulating more third-party quality signals. The question for buyers is how to use those signals coherently without creating contradictory filters that restrict scale beyond what campaign goals require. The three-tier structure of AU Media Quality - High, Average, Low - and the ability to combine segments reflects an attempt to make that calibration manageable.

Timeline

Summary

Who: Adelaide, an attention measurement company led by CEO and Co-founder Marc Guldimann, in partnership with Amazon DSP, represented by Chris Conetta, Director of Omnichannel Supply.

What: Adelaide launched AU pre-bid targeting inside Amazon DSP, introducing two products - AU Media Quality (three-tier attention quality segments for targeting or exclusion) and AU Quality Floor (a single filter excluding MFA-designated sites and the bottom 10% of AU-scored placements). Inventory coverage spans Display, Online Video, and Streaming TV across desktop, mobile web, and app environments.

When: The launch was announced on June 10, 2026.

Where: Inside Amazon DSP, accessible to buyers purchasing Display, Online Video, and Streaming TV inventory. The AU Quality Floor segment launches exclusively within Amazon DSP.

Why: The integration addresses the gap between post-campaign attention measurement and pre-bid decisioning. Advertisers previously received AU data only after a campaign ran, limiting its use to retrospective analysis. Pre-bid segments allow buyers to apply attention quality standards at the activation stage, before bidding decisions are made and budgets are committed, giving them a mechanism to reduce media waste and prioritize inventory more likely to capture attention and drive outcomes.