Commentary
AVOD Will Be Big, But Will Take Its Time Getting There
- by Dave Morgan , Featured Contributor, March 25, 2021
According to Nielsen data (crossed with second-by-second connected TV viewing data), almost 25% of TV viewing is now on streamed content. But only 4% of ad-viewing time is on streamed ads. Linear TV still represents more than 95% of total TV ad-viewing time.
This is why premium video streaming ad campaigns today are so expensive, and tend to deliver a lot of frequency and not very much reach. It’s also why ad fraud in streaming video is a real and growing problem. Ad demand exceeds supply, which is to practitioners of digital ad fraud what honey is to ants at a summer picnic.
Many folks in the media and ad industry want to change this situation. Advertisers, big tech platforms, large media and digital ad-tech companies all would love to see super-fast growth of ad-supported video on demand (AVOD).
This shift would move digital platforms into the center of the entire ad ecosystem. It would break down the separate silo that is the $65 billion U.S. TV ad industry. It would -- theoretically -- mean more yield from each ad for everyone.
However, the folks who really control whether ad-supported streaming services will take off -- viewers -- have yet to put their time and attention behind ad-supported streaming in a substantial way. And viewers are in charge of this shift, not the media industry.
The linear TV viewing shift to streaming will take many years. Older and lower-income viewers are a big part of TV viewership, are not early adopters of new technology, and face economic and structural barriers to shifting to streaming.
The latest Pew Research report tells us that 34% of U.S. households lack fixed broadband at home (that’s more than 100 million Americans), and many tens of millions of homes lack “smart” TVs. Plus, when it comes to watching live sports and news, TV’s "tent-pole" programming, broadcast and cable deliver high-definition viewing with no lag or latency.
Streamed ad experience needs massive improvement. Why does Hulu keeps giving you the same ad over and over in the same hour on the same show? How come YouTube TV leaves some ad spots blank, with dead time? How come only two major ad-supported streaming services do a human review of each and every ad they show their viewers?
Digital ad technologies may have the capacity to deliver better, more relevant ad experiences, but as we know with banners and the open web, the gap between potential and practice leaves a lot to be desired. If the industry mirrors the web ad experience on streamed TV, viewers will abandon bad-ad services in droves.
Streaming ad reporting and measurement is a mess. One of the reasons the streamed ad experience is so bad is that the streamed world is a fragmented mess. Most TV and dongle manufacturers have different operating systems. Apps on this services are different. And the ads within those services may have several owners and many paths to delivery, including: device manufacturer, operating system licenser, content owner, distributor, sell-side platforms, demand-side platforms, ad servers, creative servers, and data management platforms.
This means dozens of companies may touch each ad impression delivery, and collecting, collating and harmonizing reporting and measurement from that mess is a mess.
It’s certain that any company claiming to give advertisers an accurate, consolidated ad campaign report based on deterministic data is not telling the truth. That doesn’t exist. Broad, approximate and probabilistic is the best anyone can hope for today (and for some time).
All about time, iteration and long-term commitment. I do believe that AVOD will be a big part of the video ad ecosystem over time, but it will take years to change viewer behaviors, build a new TV-like digital ad supply and deliver on the promise of a better ad experience for advertisers, not better profits for digital ad exchange platforms that want to trade human viewers’ eyeballs like pork bellies (thank you, mentor and media legend Wenda Millard, for warning all of us about this more than a decade ago).
The winners will put viewers first, will stay committed to building ad-supported services over years and years, not quarter over quarter, and will test, learn and optimize their way to a better TV experience for all.
What do you think? Do AVOD services have to be instant successes now to be long-term winners?
Dave, you make some very good points and much of what is now called AVOD is non-premium content which major national TV advertisers aren't very interested in. However, I don't see the measurement issue as that big a problem once a substantial amount of "quality" programming---mainly by Comcast, Discovery, Viacom/CBS, and, I suspect, soon by Disney+,etc. ---is available---which will be very quickly and at reasonable scale based on their rising subscription totals. Instead of worrying about digital targeting methods, all the sellers have to do is use Nielsen ratings---GRPs----with the same audience guarantee metrics that are stilll in vogue---18-49, 25-54--- and, for the time being, operate the same way they do now in "linearTV" sales---except with higher CPMs. Later, they can get into the "granular" targeting business---if this is worthwhile and there is enough interest.
Ed, I totally agree that the premium media owners will sell a good part of their ad-supported streaming inventory on TV metrics. However, many of them are finding that they can sell at hihger CPM's in private digital marketplaces where they can get paid on all impressions (P2+, essentially) and can get some targeting premiums as well. In most cases, getting P2+ CPM's around $30.
The reporting and measurement issues though, are because there is no one stream (or set of ads) that is deliivered synchroniously and consistently to everyone, since the apps, devvices, operating systems, ad servers and exchange platforms are so fragmented. That is the part that is so messy. Right now, the media owners don't have the leverage to clean it up. More likely, it will be folks like Roku, Google and Amazon. They have much "platform power."
Dave, one thing that the TV sellers are, no doubt, counting on regarding their anticipated streaming CPMs---even if packaged and sold the usual "linear" way--- is that the demographics---younger---will favor them---especially if streaming sales are combined with linear sales in the upfront. Another reason for higher CPMs ---even without better targeting for streaming buys--- will be the much lower ad clutter rate---at least at the outset---for streaming platforms. This alone should suffice to garner a 25% or greater CPM hike for streaming.
The trick will be how to combine high streaming CPMs with lower "linear" CPMs if both are sold together in massive upfront deals. With the advertisers still using bean counting methods to judge the expertise of their time buyers, CPM---or CPP---will still be the dominant metric---which means that the buyers will resist higher CPMs for streaming. My solution: sell steraming using the same metrics but separately from "linear". But this will require a major advertiser sell---which I doubt the networks know how to execute.
Very good point Ed. For sure, of the streaming ivnentory that is being sold as or with linear inventory is skewing younger and giving them much more 18-49, if not 18-34, which will certainly get them some premium pricing on its own.
So little has changed here in the last 5-10 years. Until addressable and digital ad impressions account for more than 5% of ad impressions, resources will not be devoted to better counting -- even with better demos. But in the addressable world, demos matter less and actual buying more -- even if the buyer is over 54 they are targeted. Because the inventory is limited the same ads are served over and over (and over). Because inventory is limited pricing is higher than linear tv. Roku's recent purchase of Nielsen technology may release lots of addressable inventory, but not nearly as much as if Nielsen had stayed the course and reported on other set manufacturers. This continued fragmentation will delay scale until the manufacturers consider pooling their data or offering to a third party.
Really good points Jack. I very much agree that Nielsen must (and I hope that they haven't walked away from this yet) get data from other device companies and not just lean on Roku and devices where Gracenote code is embedded. With Roku now owning that code, it is not likelyl that there will be active attempts to distribute it standalone on other devices for industry mneasurement needs.