Netflix's Vertical Feed Puts Reels-Style Inventory Inside a $30 CTV CPM
Netflix confirmed on April 17 it is launching a TikTok-style vertical video feed in its mobile app at the end of April 2026, paired with GenAI-based recommendations. That feed arrives as Netflix chases a $3 billion 2026 ad target and readies Q2 interactive ad formats, creating inventory that looks less like prestige CTV and more like Reels. Media buyers pricing Netflix as one flat premium line item should separate feed CPMs before blended rates lock.
The product change is smaller than the story. The ad change is bigger.
On paper, a vertical feed for show discovery sounds like a feature update. You swipe, you see a clip, you tap to watch the full title or save it to your list. Netflix first tested this on iPhone in May 2025 and kept iterating before rolling it wider.
The more interesting part is what Netflix said the same week. Co-CEO Gregory Peters told investors there is still "tremendous room" to improve personalization with newer technologies. Ted Sarandos framed the InterPositive acquisition as accelerating generative AI "specifically for filmmakers." Both quotes are about content. But Netflix also said explicitly it wants to use AI to improve its ad suite and allow "new formats and customization to get better returns."
That's the sentence that matters to anyone running Netflix buys. A vertical feed means mid-scroll ad slots. AI-driven ad formats mean those slots will be programmatic, short, and probably skip-tolerant. Which is a very different product than a 30-second pod inside a binge-watch, and it's about to be sold under the same Netflix SKU.
Why the CPM math is about to get weird
Netflix inventory averages roughly a $30 CPM, with premium campaigns landing between $25 and $65. Programmatic already moves close to half of the non-live business. More than 60% of new subscribers in markets where the ad tier is available choose it, and the ad-supported base hit 190 million monthly active viewers globally as of late 2025.
Those CPMs were priced on a specific product. Lean-back viewing, high attention, limited ad load, co-viewing context, a user who had already committed to finishing the episode. The vertical feed breaks most of those assumptions. Feed viewing is lean-forward and low-attention. Ad load scales with scroll speed, not episode length. There is no co-viewing. The viewer hasn't committed to anything past the next swipe.
From what I've seen priced on Reels, comparable short-form feed CPMs tend to sit in the $8 to $15 range once you strip out influencer-driven buys. If Netflix runs one unified price across long-form and feed inventory without breaking it out, advertisers will effectively subsidize cheap impressions with premium dollars. If Netflix prices them separately, it dilutes the prestige story it has been selling the Street. Either way, this is a media planning problem that does not exist today and will exist by the Q2 interactive ads launch.
There is a second-order issue the CPM conversation misses. Measurement on premium CTV is already uneven, and the tools most buyers use (Nielsen, iSpot, first-party panel-based measurement) were built around sit-down viewing, not scroll depth. A vertical feed introduces a metric that nobody has standardized on the buy side: completion at 3 seconds, 5 seconds, view-through under 50% duration. Those are Reels and Shorts metrics. Whether Netflix reports them, rolls them into the same dashboard as long-form completion rates, or lets them flow through the same programmatic pipes without flagging is a detail that will quietly decide whether this feed inventory becomes a bargain or a landmine for the next 18 months.
The question to put in writing before late April
Netflix's ads leadership has been signaling what's coming. Interactive formats ship in Q2 2026. Pause ads are in testing. First-party data is getting "more accessible" for measurement. The advertiser base grew to more than 4,000 last year, up 70% year over year. A lot of those are new clients who signed up expecting prestige CTV and are about to be offered a format that does not look like it.
The practical question for any agency with a real Netflix line item is whether you can buy feed and long-form inventory separately, with separate CPMs, separate creative specs, and separate reporting. Honestly, nobody knows the answer yet, because Netflix hasn't published a rate card for inventory that launches in two weeks. I'd push your Netflix rep on this specific question now, in writing, before the feed goes live. Once it's in-market and rolled into programmatic, you're buying a blended CPM and calling it strategy.
The content model has an adjacency problem too
The GenAI personalization change is the other half of the story, and it has been getting less attention. Netflix is pushing AI recommendations "using modern model architectures," which in practice means the feed you see won't just be clips it thinks you want. It will be clips the model thinks will extend your session. Those are not the same objective. Session-extension algorithms optimize for stickiness, not quality. Instagram and TikTok spent three years explaining why that distinction matters to brands worried about context.
There is also a brand-safety gap that nobody is really discussing. If the vertical feed eventually mixes in video podcasts as reported, the adjacency conversation shifts. Podcast episodes aren't produced like originals. They're conversational, live-tracked, occasionally off the cuff. A brand buying Netflix because it thought it was buying Stranger Things adjacency might wind up running next to a podcaster riffing on politics. IAS and DoubleVerify don't have great tooling for this kind of context yet, and the press cycle on "brand ran next to podcast X" feels like a 2026 story waiting to happen.
The line-item audit to run before Netflix flips the switch
If Netflix is more than 5% of your ad mix, run three numbers before the end of April. First, pull your blended Netflix CPM for the last two quarters so you have a clean baseline. Second, ask your Netflix rep for a written commitment on how the vertical feed inventory will be priced and reported in Q2. Third, reprice your 2026 Netflix plan assuming a 30% CPM dilution from feed inventory. That is roughly the gap between premium CTV and short-form feed on other platforms. It isn't a precise number, it's a reasonable worst case to plan around.
While you have the rep on the phone, there are two other asks worth putting in writing. One: can you opt out of feed inventory entirely for a given campaign, or is it bundled by default once Netflix's AI starts allocating based on what it thinks converts for your brand? Two: will creative specs differ between long-form pods and feed placements, and if so, who bears the production cost of reshooting for a vertical aspect ratio? Neither question has a clean answer today, which is exactly why it's worth asking now rather than after the launch.
A lot of marketers have been treating Netflix ads as a high-trust line item because the product it was attached to felt premium. That trust was priced in. Netflix's $3 billion ad target was already stretching that trust before Reed Hastings exited the board. A vertical feed with AI-curated clips inside the same SKU is the moment the line item starts meaning something different.
I don't think most buyers will push Netflix on this in time. The feed will ship, the blended CPM will stick, and the reporting will look fine on paper while a chunk of spend quietly moves into inventory nobody quoted. Worth making noise now, while Netflix still has a reason to answer.
Notice Me Senpai Editorial
