Roughly 26% of programmatic CTV impressions still get bought without any real signal on whether the ad even rendered, according to industry measurement estimates circulating through 2025 into this year. That’s the waste XR ONE claims to solve with its new AI format-prediction layer. The pitch: let a model decide whether a given impression should run as CTV or standard digital display, before the bid ever clears. Does it work, or is this another dashboard dressed up as intelligence?
What the Format-Prediction Layer Actually Does
XR ONE’s format-prediction layer sits between the DSP bid request and the creative decisioning engine. Instead of a media buyer manually splitting budget across CTV and digital line items, the model ingests device signals, historical completion rates, and inventory quality scores, then recommends — in real time — which format is more likely to convert for that specific impression opportunity.
In theory, this is a smarter version of what trading desks have done manually for years: shift dollars toward whichever channel is outperforming. The difference is speed and granularity. A human planner reallocates budget weekly, maybe daily if they’re diligent. XR ONE’s layer claims to make that call at the impression level, thousands of times per second.
That’s a meaningful architectural shift. It’s also exactly the kind of claim that needs scrutiny before finance signs off on a bigger commitment. We covered the platform’s broader ad-ops consolidation play in our unified ad-ops platforms buyers guide, and the format layer is the newest, most aggressive piece of that stack.
The Waste Problem It’s Supposedly Solving
CTV waste isn’t a new complaint. Brands have been grumbling about it since the format went mainstream. The usual suspects: ad stacking on connected devices, bots inflating impression counts, and — the one nobody likes to admit — genuinely mismatched creative running on the wrong screen size for the wrong context.
eMarketer’s audience research has repeatedly flagged that CTV ad load and completion behavior vary wildly by app and device manufacturer, which means a blanket CTV strategy is almost always going to burn budget somewhere. Digital display has its own waste profile: viewability issues, bot traffic, and banner blindness that’s been documented since display advertising’s early days.
The real question isn’t whether CTV or digital wastes more money individually — it’s whether an automated layer can outperform a skilled human planner at deciding which format fits which impression, in real time, without introducing new blind spots.
That’s a much harder bar to clear than “reduces waste.” Reducing waste compared to doing nothing is trivial. Reducing waste compared to a well-run manual or semi-automated program is the actual test.
How the Model Makes Its Call
XR ONE’s documentation (and conversations with agency partners testing the beta) suggest the prediction layer weighs four signals: device/browser environment, time-of-day engagement patterns, historical format performance by advertiser vertical, and a proprietary “attention likelihood” score built from partial-view and completion data.
That attention score is the interesting part, and also the part with the least transparency. XR ONE won’t publish the model architecture, which is standard practice for competitive reasons but makes independent validation nearly impossible. If you’ve read our piece on evaluating AI format-matching tools, you already know this is the recurring problem across the category: vendors ask you to trust the output without showing the inputs.
Practically, this means brands adopting the format-prediction layer are trusting a black box to reallocate real budget. That’s not disqualifying — plenty of effective marketing tech operates this way — but it does mean your evaluation criteria need to shift from “does the model make sense” to “does the output hold up under audit.”
Does It Actually Cut Waste? What the Numbers Suggest
Early adopter data is thin, and XR ONE hasn’t released independently verified benchmarks. But agencies running parallel tests (XR ONE’s layer active on half of a campaign’s budget, manual allocation on the other half) report format-prediction impressions delivering completion rates 8-14% higher than manually split budgets, with a corresponding reduction in reported invalid traffic.
That’s a promising early signal. It’s also a small, self-selected sample from agencies that opted into a beta program, which is not the same as a rigorous, third-party controlled study. Treat it as directional, not conclusive.
The bigger structural benefit might not be the waste reduction itself but the audit trail it creates. Every format decision gets logged with the signals that drove it, which matters enormously if you’re building the kind of AI governance documentation regulators and internal compliance teams increasingly expect. Our AI governance scorecard for vendors is a useful companion checklist here — format-prediction tools should be evaluated on explainability, not just performance lift.
Compare this to the earlier waste study we ran on XR ONE’s broader platform. In our prior waste-reduction analysis, the consolidated ad-ops layer showed clearer, more measurable savings on operational overhead (fewer platform fees, less duplicate tagging) than on media waste specifically. The format-prediction layer is a narrower, riskier bet: it’s asking the model to make a judgment call that used to require human context.
Where This Breaks: Attribution and Identity Gaps
Here’s the part vendors gloss over. Format-prediction is only as good as the identity signal feeding it. CTV environments are notoriously fragmented when it comes to cross-device identity — a household might have three streaming devices, two smart TVs, and a laptop, and stitching those into a single audience view is still an unsolved problem industry-wide.
If XR ONE’s model is predicting format based on incomplete identity resolution, it’s making confident recommendations on shaky data. This isn’t unique to XR ONE — it’s the same identity fragmentation problem we’ve flagged repeatedly, including in our breakdown of fixing identity fragmentation before it breaks AI narratives. Any AI layer making channel decisions is inheriting the identity quality of whatever graph sits underneath it.
Brands running XR ONE’s format layer alongside a strong identity resolution partner (LiveRamp, Epsilon, Acxiom — see our identity resolution buyers guide for a full comparison) will get materially better predictions than brands running it on cookie-adjacent, low-confidence signals. That’s not a knock on XR ONE specifically. It’s a reminder that no format-prediction layer, however sophisticated, can outrun bad upstream data.
The Attribution Question Nobody’s Answering Yet
Say the model shifts 30% of your budget from digital to CTV mid-flight. Great — but how do you attribute the resulting sales lift? Standard last-touch attribution wasn’t built for impression-level format switching happening in milliseconds. You need a measurement stack that can handle granular, real-time channel shifts without losing the thread on what actually drove conversion.
This is where a lot of brands underestimate the lift required to actually operationalize a tool like this. It’s not just “turn on the feature.” You need a data warehouse capable of ingesting format-level bid decisions alongside conversion data — something we’ve discussed at length in our piece on why marketing attribution needs a warehouse. Without that infrastructure, you’re trusting XR ONE’s self-reported performance dashboards, which is a conflict of interest baked right into the product.
Also worth asking: who’s monitoring the model itself? AI systems drift. A format-prediction layer trained on one quarter’s device and app inventory mix can degrade quietly as the CTV app ecosystem shifts (and it shifts constantly — new ad-supported tiers, new bundling deals, new device manufacturers entering the space). If you don’t have monitoring in place for marketing AI agents, you won’t catch that degradation until the waste creeps back.
A Practical Framework Before You Buy In
If you’re evaluating XR ONE’s format-prediction layer, or any competitor building similar CTV-vs-digital automation, run it through these checks before committing real budget:
- Demand a holdout test. Run the model on a defined budget slice with a manually-managed control group running in parallel, for at least one full sales cycle.
- Audit the identity input. Ask what identity graph feeds the model and how confident the cross-device matches are for your specific audience.
- Check the attribution stack compatibility. Confirm your measurement infrastructure can capture impression-level format decisions, not just channel-level totals.
- Get the drift-monitoring plan in writing. Ask how often the model retrains and what triggers a retrain outside the normal schedule.
- Verify cost governance. Real-time inference at impression scale has compute costs. Understand how those get passed through, and read our FinOps guide to marketing AI compute spend before signing anything with variable pricing.
None of this is exotic due diligence. It’s the same rigor you’d apply to any vendor claiming to automate a six- or seven-figure budget decision. The fact that it’s wrapped in “AI” doesn’t lower the bar — if anything, it raises it, because the decision logic is harder to inspect than a spreadsheet formula.
For broader context on where AI marketing tools tend to overpromise, eMarketer’s research on ad tech adoption and Statista’s CTV ad spend data are both useful benchmarks for sanity-checking vendor claims against actual market trends. The FTC’s guidance on automated decision-making disclosure is also increasingly relevant as these tools make budget calls with less human oversight.
FAQs
Frequently Asked Questions
Does XR ONE’s format-prediction layer actually reduce ad waste?
Early agency-reported data shows completion rates 8-14% higher on impressions where the model chose the format, compared to manually split budgets. That’s a promising early signal, but it comes from a small, self-selected beta sample rather than an independent, controlled study, so treat the numbers as directional rather than conclusive.
How does XR ONE decide between CTV and digital display for a given impression?
The model weighs device and browser environment, time-of-day engagement patterns, historical performance by advertiser vertical, and a proprietary attention-likelihood score built from partial-view and completion data. The full model architecture isn’t public, so independent validation is limited.
What’s the biggest risk in adopting an AI format-prediction tool for CTV budgets?
Identity fragmentation. If the underlying identity graph has weak cross-device matching, the model is making confident format calls on incomplete data. Attribution is the second major risk, since standard measurement models weren’t built to track impression-level format switching in real time.
Is XR ONE’s format layer different from its ad-ops consolidation platform?
Yes. The unified ad-ops platform focuses on operational consolidation, reducing platform fees and duplicate tagging. The format-prediction layer is a narrower, newer feature that makes channel-level bidding decisions using AI, which carries a different (and arguably higher) risk profile.
What should brands require before committing budget to this kind of tool?
A parallel holdout test against manual allocation, transparency on the identity data feeding the model, confirmation that existing attribution infrastructure can capture impression-level decisions, and a written plan for model drift monitoring and retraining.
The bottom line: XR ONE’s format-prediction layer shows real early promise on completion rates, but “cuts waste” is a claim that only holds up against a rigorous holdout test, not a vendor dashboard. Run the parallel test before you reallocate a single working dollar.
Frequently Asked Questions
Does XR ONE’s format-prediction layer actually reduce ad waste?
Early agency-reported data shows completion rates 8-14% higher on impressions where the model chose the format, compared to manually split budgets. That’s a promising early signal, but it comes from a small, self-selected beta sample rather than an independent, controlled study, so treat the numbers as directional rather than conclusive.
How does XR ONE decide between CTV and digital display for a given impression?
The model weighs device and browser environment, time-of-day engagement patterns, historical performance by advertiser vertical, and a proprietary attention-likelihood score built from partial-view and completion data. The full model architecture isn’t public, so independent validation is limited.
What’s the biggest risk in adopting an AI format-prediction tool for CTV budgets?
Identity fragmentation. If the underlying identity graph has weak cross-device matching, the model is making confident format calls on incomplete data. Attribution is the second major risk, since standard measurement models weren’t built to track impression-level format switching in real time.
Is XR ONE’s format layer different from its ad-ops consolidation platform?
Yes. The unified ad-ops platform focuses on operational consolidation, reducing platform fees and duplicate tagging. The format-prediction layer is a narrower, newer feature that makes channel-level bidding decisions using AI, which carries a different (and arguably higher) risk profile.
What should brands require before committing budget to this kind of tool?
A parallel holdout test against manual allocation, transparency on the identity data feeding the model, confirmation that existing attribution infrastructure can capture impression-level decisions, and a written plan for model drift monitoring and retraining.
Top Influencer Marketing Agencies
The leading agencies shaping influencer marketing in 2026
Agencies ranked by campaign performance, client diversity, platform expertise, proven ROI, industry recognition, and client satisfaction. Assessed through verified case studies, reviews, and industry consultations.
Moburst
-
2
The Shelf
Boutique Beauty & Lifestyle Influencer AgencyA data-driven boutique agency specializing exclusively in beauty, wellness, and lifestyle influencer campaigns on Instagram and TikTok. Best for brands already focused on the beauty/personal care space that need curated, aesthetic-driven content.Clients: Pepsi, The Honest Company, Hims, Elf Cosmetics, Pure LeafVisit The Shelf → -
3
Audiencly
Niche Gaming & Esports Influencer AgencyA specialized agency focused exclusively on gaming and esports creators on YouTube, Twitch, and TikTok. Ideal if your campaign is 100% gaming-focused — from game launches to hardware and esports events.Clients: Epic Games, NordVPN, Ubisoft, Wargaming, Tencent GamesVisit Audiencly → -
4
Viral Nation
Global Influencer Marketing & Talent AgencyA dual talent management and marketing agency with proprietary brand safety tools and a global creator network spanning nano-influencers to celebrities across all major platforms.Clients: Meta, Activision Blizzard, Energizer, Aston Martin, WalmartVisit Viral Nation → -
5
The Influencer Marketing Factory
TikTok, Instagram & YouTube CampaignsA full-service agency with strong TikTok expertise, offering end-to-end campaign management from influencer discovery through performance reporting with a focus on platform-native content.Clients: Google, Snapchat, Universal Music, Bumble, YelpVisit TIMF → -
6
NeoReach
Enterprise Analytics & Influencer CampaignsAn enterprise-focused agency combining managed campaigns with a powerful self-service data platform for influencer search, audience analytics, and attribution modeling.Clients: Amazon, Airbnb, Netflix, Honda, The New York TimesVisit NeoReach → -
7
Ubiquitous
Creator-First Marketing PlatformA tech-driven platform combining self-service tools with managed campaign options, emphasizing speed and scalability for brands managing multiple influencer relationships.Clients: Lyft, Disney, Target, American Eagle, NetflixVisit Ubiquitous → -
8
Obviously
Scalable Enterprise Influencer CampaignsA tech-enabled agency built for high-volume campaigns, coordinating hundreds of creators simultaneously with end-to-end logistics, content rights management, and product seeding.Clients: Google, Ulta Beauty, Converse, AmazonVisit Obviously →