Netflix has an opportunity to do well in the connected TV (CTV) space in terms of its upcoming advertising CTV option if it can get to a $25 to $45 cost per thousand viewer pricing, according to an assessment from Standard Media Index.
SMI says Netflix has a clear opportunity to hit this "sweet spot" -- an area that few other connected TV/digital media sellers occupy.
Looking at average CPMs from June 2021 to May 2022, SMI says that Hulu is just below this mark at $24 CPM in terms of average CPM. At the high end is HBO Max -- its advertising-supported option -- at more than $50.
>Only two services fall in between these numbers: Peacock at near $46 CPM and discovery+ at $28.
SMI says Roku comes in at around $23; Tubi, $21; and Pluto, $17. For Peacock and HBO Max, SMI says exclusive content allows for higher CPMs compared to other services.
At the low end is YouTube, at $16, while its more premium curated content -- called YouTube Select -- is at $18.
However, YouTube is the current leader in median monthly impressions of around 1.7 million, followed by Discovery+ at 1.4 million; Roku with 900,000, and Hulu and Peacock, each with 700,000. HBO Max is at around 250,000.
By way of a proxy for Netflix, SMI also included Microsoft's Xbox Network in its results -- with a $23 average CPM, and 1.1 million median monthly impressions.
Netflix recently struck an advertising partnership deal with Microsoft.
Standard Media Index (SMI) captures actual invoicing data from all major holding companies and most major independents, representing 95% of national brand ad spend.
Interesting data Wayne. Two questions. First, are these CPMs for all viewers aged 2+? And, second, are they for 30-second units or all message lengths averaged out?I assume that SMI will be willing to clarify on these questions--or, maybe, their report/press release has the info.
A couple of answers to your questions (from SMI):
-- Are these CPMs for all viewers aged 2+?
-- Are they all for 30-second units or all message lengths averaged out?
SMI: Digital CPM reflects a blended average CPM. It is a combination of various ad / placement attributes, thus it reflects a rolled-up view across the referenced dimensions.
Thanks, Wayne, that's a big help.
It will be interesting to see how Nielsen reports on individual program episodes if it tries to supply "audience "estimates for each commercial. Will we get an average commercial rating per show? No, that won.t work as the number of commercials and the mix of lengths will vary so you have no comparability. Will we get averages for commercial lengths by show? Again, this is problematic for the same reason. More important is this question, "Will we get a true picture of how many people actually "watched" each commercial"?. Nope. All we will get is a projection of how many "might have watched"---including that 30-35% who left the room. and counting all of those who stayed but paid no attention---in short,"impressions".
Very interesting Wayne and Ed.
Ed you raise very good and valid issues about how ads are reported as averages? Average across the program content? Average across the ad-break? Average across the seconds of each individual ad? They are all technically possible in the current TV system. Maybe they should all be available ... horses for courses.
I agree that it will be "might have watched" rather than "actually watched". Would "actually watched" need a threshold? Maybe six seconds in a 30? Would that be nett duration across the 30 or would it be at least six consecitive seconds? And that still doesn't address the issue of "attention" or "impact" etc.
One of the delicious ironies is that the funding model is deficient.
The network/broadcaster/streamer measure THEIR content. They then do calculations in order to work out how much an ad should cost - i.e. CPM. In fact they are multiple calculations to account for demographics.
Meanwhile the advertisers are pushing for the network to extend their existing research system (if possible) to measure the key metrics for their ads - opportunity to view, viewed, attentiveness, impact etc. But I wonder how much they are willing to pay to get that research done on their behalf? My guess would be 4x the CPM. Worth thinking about eh.