Meta's Creator Ads Hit $15B, and They Cost More per Impression, Not Less
Meta said on its October 21 Q3 2026 earnings that partnership ads (ads that run through a creator's handle) grew over 97% year over year to a $15 billion annual revenue run rate. In the same quarter, average price per ad rose 23% in the US and Canada. Partnership ads tend to cost more per impression, so judge them on cost per incremental purchase, not CPM.
That last point matters because the easy reading of this news is "creator ads are the cheap inventory, move budget there before Black Friday." I don't think the data supports that, at least not the CPM part. What it supports is something a bit narrower and, honestly, more useful.
The $15 billion number, with the asterisks attached
First, the figure itself. According to PPC Land's write-up of the earnings call, Meta told analysts partnership ads grew "over 97% year-over-year" and "surpassed $15 billion in annual revenue run rate." Back in Q1, Zuckerberg put the run rate at $10 billion, as PPC Land reported at the time. So that's roughly a 50% jump in six months.
Two labels worth keeping straight. A run rate is an annualized snapshot (usually the latest quarter or month multiplied out), not reported revenue. And Meta doesn't break partnership ads out as a line item in its filings, so there's no way to audit the $15 billion. It's a number Meta chose to say out loud on a call. For context, total Q3 ad revenue was about $63 billion for the quarter, so even taken at face value, creator-handle ads are a small slice of a very large pie.
Still, doubling in a year while the core business grew 26% means advertisers are shifting money into this format faster than into almost anything else Meta sells. That part seems real.
Where the 23% price jump actually landed
The other half of the story is pricing. Meta's regional breakdown shows worldwide ad impressions up 10% and average price per ad up 14%. In US and Canada, impressions rose 8% while price per ad rose 23%. Europe was +10% on price. Asia-Pacific was basically flat at +1%.
Quick label check: "average price per ad" is Meta's blended metric across every auction, objective and placement. It isn't your CPM. But when the blended number climbs 23% in your biggest market, your CPMs are very likely climbing too, and PPC Land notes that the US price increase has now accelerated three quarters running, from 14% to 20% to 23%.
Meta guided Q4 revenue to $73 billion to $76 billion. Q4 is when the auction gets crowded anyway. Stack that on top of a market where prices are already rising faster than impressions, and you get a holiday season where the cost of reach probably goes up more than most media plans assumed back in August.
Partnership ads are not the cheap inventory
This is the part the "move budget to creators" crowd skips. The best public dataset I've found on this comes from Agentio, a company that sells creator-ad tooling (so, keep that in mind). Its report, covered by Net Influencer in August, looked at $130 million in spend across 65,000 ads from 137 brands. Partnership ads ran at a 19% higher CPM than the same kind of licensed creator content run from the brand's own handle.
And yet they still came out ahead where it counts: 19% higher CTR, 10% higher conversion rate, 5% lower CPA.
Placement mattered a lot. In Explore, partnership ads had a 57% lower CPM and 129% higher CTR. In Stories, they lost on CPA by 25%. So "partnership ads cost more" isn't even uniformly true. It's an average that hides some pretty big swings in both directions.
On paper, paying 19% more per thousand impressions sounds like the wrong move during a price spike. And sometimes it is. But a 5% better CPA on a 19% pricier impression suggests the creator's handle is doing conversion work the brand handle can't. It just doesn't do it for free.
The question for Q4 is which format survives a 23% price hike with its CPA intact. CPM won't tell you that.
Meta's own 45% claim, read carefully
Meta has been pushing a specific allocation number. Its creator marketing white paper from October 7 says advertisers who put at least 30% partnership ads into business-as-usual campaigns saw a 45% lower median cost per incremental purchase, based on 20 EMEA conversion lift studies. PPC Land's breakdown points out the meta-analysis ran at 92% confidence, below the usual 95% bar, and the studies date from the first half of 2025.
Twenty studies from one region, well over a year old, from the company selling the inventory. I'd treat 30% as a reasonable starting test cell rather than a proven allocation. Meta has a habit of publishing lift numbers without a clean comparison set (we picked apart its 18.9% GEM conversion claim last week for the same reason).
And to be fair, this isn't entirely new. Brands have been running whitelisted creator content for years. What changed is that Meta built the plumbing (the Creator Marketing Hub, which Meta said is now rolled out to all advertisers with over 10 million creators discoverable) and is now actively steering budget toward it. When a platform makes one format easier to buy and talks it up on an earnings call, the auction for that format tends to get more crowded. Which, if you think about it, is exactly how CPMs go up.
A two-week test before the holiday bids lock
Here's what I'd actually run, starting this week, if I had a Meta account spending real money in the US:
- Pull your last 30 days by identity. Split ads into brand-handle and partnership (creator-handle) and look at CPM, CTR and CPA side by side. If you've never separated them, you probably don't know which one is carrying your account.
- Break it out by placement. Agentio's data suggests Explore and Feed behave very differently from Stories. If your partnership ads mostly run in Stories, the format may be losing you money and the account-level average is hiding it.
- Set a CPA ceiling, not a CPM ceiling. A rough benchmark: if partnership ads carry up to about 20% higher CPM than brand ads and still land at or below brand CPA, they're earning their premium. If CPM is 30% or more higher and CPA is worse, cut back.
- Run a lift test at the 30% mark. Put roughly 30% of one campaign's assets into partnership ads and run a conversion lift study against a control. Two weeks is tight, but it gets you a directional read before BFCM budgets are set.
- Price the creator fee in. A partnership ad that beats brand CPA by 5% but carries a $3,000 usage fee may not beat it at all. Cal AI's approach of capping creator CPMs against its own revenue per thousand views is a decent model for that math.
None of this is complicated. It's mostly a reporting split most teams haven't bothered to make, and Meta's reporting doesn't exactly make it obvious, either.
My guess on where creator CPMs go next
I think partnership ads cross a $20 billion run rate by the Q1 2027 call, and I'd bet the CPM premium over brand-handle ads widens, not narrows, as more advertisers pile into the same pool of high-performing creators. Ruggable moving $1 million from Meta to ShopMy creators is one example of brands already treating creators as a media line rather than a content line, and that tends to bring media-style price inflation with it.
That doesn't make the format a bad buy. It just means the cheap-reach window, if there ever was one, is probably mostly gone, and the advantage left is conversion quality. Which is a fine reason to buy something. It just needs to be the reason you're actually measuring.
From what I've seen, the accounts that do well in Q4 aren't usually the ones that found a magic format. They're the ones that knew, before the auction got expensive, which of their ads were worth paying more for.
