Why audio CPMs outperform display in 2026 — and where the ceiling is
Ask any trader who buys both formats: the same budget, moved from display banners into streaming audio, tends to come back with better completion, better recall and — increasingly — better cost per action. That isn't a slogan; it's a structural property of the medium. It's also not unlimited. Here is how we think about it from inside the exchange.
Attention you can't scroll past
A display impression competes with everything else on the page. An audio impression is the page. When a listener is streaming a podcast or live radio, the ad occupies the entire sensory channel for its duration — there is no banner blindness, no tab in the background, no scroll. That is why audio completion rates routinely land near 90% while display viewability still hovers around the industry's much lower bar.
Completion is the currency here. An advertiser buying a 30-second audio spot generally gets 27+ seconds of actual exposure. A display buyer counting "viewable impressions" is often paying for one second of 50% pixels. When you normalize price per second of real attention, audio's apparent CPM premium usually inverts into a discount.
The measurement gap is closing
Audio's historical weakness was attribution: no click, no pixel, no easy conversion story. That has changed on three fronts:
- Quartile and audibility reporting is now standard on well-implemented DAAST/VAST audio, so buyers can verify exposure instead of assuming it.
- Promo-code and QR-style attribution matured from podcast advertising into programmatic audio generally.
- Enriched bid requests — content genre, listening context, geo — let buyers target and measure against real signals instead of proxies.
Each of these shifts moves audio spend from "brand experiment" to "performance line item," and performance budgets are what push CPMs up sustainably.
Supply is still under-priced — for now
Podcast and streaming inventory grew faster than the demand infrastructure that buys it. Plenty of quality audio supply still trades below the attention it delivers because buyers can't yet evaluate it properly — the metadata is thin, the supply path is murky, or the seller isn't discoverable. That's the arbitrage window. Publishers who fix their metadata and their supply-chain declarations (see our supply chain piece) capture the gap; those who don't leave money with intermediaries.
Where the ceiling is
Honesty matters here: audio CPMs do not rise forever, and three forces cap them.
- Frequency tolerance. Listeners tolerate fewer ads per hour than a webpage tolerates banners. Ad load discipline is a feature of the medium, and it caps sellable impressions.
- Creative supply. Great audio creative is scarcer than great display creative. Markets where every buyer runs the same three text-to-speech reads will plateau early.
- Contextual limits. Audio is heard, not seen — product categories that depend on visual demonstration will always allocate elsewhere first.
The realistic bull case for audio isn't "CPMs to the moon." It's audio taking its proportional share of attention budgets — and attention is the scarcest asset in advertising.
What we'd tell each side
Advertisers: normalize your channel comparison to cost per attentive second, not CPM. Audio will surprise you, and you'll be early rather than late to properly-priced inventory.
Publishers: your ceiling is set by metadata quality and supply-path cleanliness more than by content quality. Fix the plumbing; the price follows.