A Guide to Ad Inventory 2026: What Still Works, What's Changed

Ad inventory hasn't completely changed, but it has picked up new features. The underlying formats are largely the same, and so are the metrics.

What's changed is where the money flows, which paths get trusted most, and which ones are quietly getting cut off.

This guide will help you make sense of the constants and the shifts in ad inventory, so you can navigate the landscape in 2026 and beyond.

‍

What Still Works

Premium and Remnant Still Split the Market — Just Differently

Premium is still what you know it to be: above-the-fold, homepage, exclusive, sold directly or through private marketplaces. Remnant is still whatever's left, cleared through the open exchange. That split hasn't gone anywhere.

What's moved is the label. Remnant no longer means cheap or disposable. On a properly selected supply path, remnant impressions can still reach engaged audiences. In 2026, remnant is a transaction type, not a quality indicator.

That means you should stop discounting inventory just because it's labeled remnant. Instead, ask which supply path it traveled to reach you.

‍

Impressions, Fill Rate, eCPM — Still the Most Crucial Metrics

The three metrics that define ad inventory remain unchanged: impressions, fill rate, and eCPM. Impressions are the unit. Fill rate tells you how much available inventory actually sold. eCPM tells you what you're paying per thousand.

None of them work in isolation. Impressions without viewability just count ads no one saw. Fill rate without price tells you demand exists, not whether it's worth meeting. eCPM without volume can hide a minuscule, cherry-picked sample dressed up as a trend.

The discipline hasn't changed because it didn't need to. Buyers who triangulate all three still catch what buyers chasing one metric miss.

‍

Five Pricing Models, No New Ones

The pricing models of yesterday are the pricing models of today. CPM, vCPM, CPC, CPA, and flat-rate are still the staples — no sixth model has replaced them. Automated bidding layers like CPA Goal and SmartCPM sit on top of these five; they don't add a new way to pay.

Choosing a pricing model was never about which one is best. It's about which one carries the most risk if the impression doesn't perform. CPM puts the risk on the advertiser; CPA puts it on the publisher or network. The other models sit somewhere in between. That risk calculus hasn't changed.

‍

Contextual Targeting Didn't Adapt to Cookieless. It Was Already Built For It.

Contextual targeting hasn't changed because it never depended on the thing that did. It matches ads to content, not people — so ATT, cookie deprecation, and tightening privacy regulations never touched its inputs. Every other targeting method had to rebuild around identity loss. Contextual just kept running, unchanged, the whole time.

That's why 2026 treats it as the default, not the fallback. It's the one targeting method that works consistently whether you're buying on the open programmatic exchange or inside a curated network.

‍

What's Changed

The Open Exchange Is the New Remnant

The hierarchy has flipped. More than 91% of U.S. programmatic display ad spend now runs through private marketplaces and programmatic direct deals. The open exchange, once the default way to buy programmatic, is now the smaller, secondary option.

This isn't buyers being more cautious for its own sake. It's buyers paying for quality, brand safety, and measurement transparency that the open exchange has never reliably delivered. The premium advertisers pay for PMP and direct inventory is the cost of trust, not a markup.

If your buying strategy still treats the open exchange as the main route and PMPs as the side street, it's worth reconsidering — the model has inverted.

‍

Four in Ten Display Auctions Are Still Taking the Scenic Route

Rebroadcasting — the practice of reselling inventory through a chain without exclusive rights to sell it — accounted for 41% of display auctions and 26% of video auctions as of April 2026. It's not a shrinking problem. It's a persistent one, and it's actually grown in display even as it's declined in video.

It's also harder to spot than it used to be. Self-declared transparency labels can't be taken at face value anymore: 15% of bid requests independently identified as wasteful rebroadcasting are still labeled DIRECT in publisher ads.txt files. The paperwork says one thing; the actual payment chain says another.

The market has responded accordingly. The Trade Desk has moved to a near-comprehensive platform-wide block on rebroadcasting, and other DSPs are following. That's not a routine policy update — it's a sign that rebroadcasting is now seen as a disqualifying practice.

‍

Rebroadcasting Now Costs You Half Your CPM

Here's a number that should get your attention: rebroadcasting supply chains carry a 50% monetization deficit compared to maximally direct paths. They're structurally worth half as much per bid request, and DSPs are already pricing them that way.

That's what makes supply path optimization non-negotiable in 2026. If your DSP can't tell you how many hops sit between your spend and the publisher, you may be funding that 50% gap without knowing it. The fix is to audit the payment chain, not just the ads.txt labeling.

‍

New Surfaces, Same Old Rules

Telegram Mini-App Ads, Alexa+ Agentic Ads, and Google Maps ad targeting didn't exist as scaled inventory just two years ago. Today, they're real, buyable surfaces.

But you can't treat them like the formats you already know. They behave differently, which means separate budgets, longer evaluation windows, and a tolerance for noisier early data than you'd accept from display or CTV.

What hasn't changed is the underlying evaluation framework. Quality still gets checked the same way. Targeting still gets layered the same way. Measurement still has to prove out before you scale. New surface, same diligence.

‍

Moving at a Brisk Pace

AI, both generative and agentic, is changing the ad industry at a rapid pace. Teams can generate ad creative, from copy to visuals to messaging, in bulk and nearly instantaneously. They can also set media buying strategy while agents handle targeting and bidding.

Ad inventory is changing too, just more slowly. There have been real shifts, but most of the fundamentals have held steady.

That calls for a two-pronged approach. First, double down on the basics: measuring the right metrics, optimizing your supply paths, and choosing pricing models based on risk. Second, recognize what's new: PMP dominance, the real cost of rebroadcasting, and an expanding inventory ecosystem.

Balance the old and the new, and your ad inventory purchases will be safer, more cost-efficient,  and better aligned with your budget.

Are you looking to maximize your ad inventory strategy? Get in touch with us to learn how we can optimize your supply path, targeting and more for maximum performance. 

‍

Frequently Asked Questions

What are the best programmatic ad platforms?

The best programmatic ad platform for your team and clients ultimately depends on your marketing objectives and requirements. However, the following programmatic ad platforms are known across the board for their offerings and support:

• Facebook Ads
• LinkedIn Ads
• TheTradeDesk
• Simpli.fi

How to measure programmatic ad success?

There are numerous metrics advertisers can measure to determine success, and the right ones will ultimately depend on their marketing goals. That said, seven key metrics include:

• Impressions
• Clicks/Click-through rate (CTR)
• Conversion rates
• Cost
• Revenue
• Reach
• Return on ad spend (ROAS)

What is scale in programmatic advertising?

In the context of programmatic advertising, scale would revolve around a DSP’s ability to reach as many eligible consumers as possible. The industry offers a standardized form of measurement to help determine scale known as QPS, or queries per second.

‍