Total US TV (Incl. Streaming/CTV) Ad Time To Fall, Estimated 24% Decline Between 2023 and 2027

Beginning in January 2024, most Madison and Wall research and related data published on this Substack will primarily be available to consulting or advisory clients, or otherwise as part of a paid corporate subscription.   Please reach out to brian@madisonandwall.com if you would like to discuss these services or for more information about the new offering.  

What are the consequences of the growth of streaming services on TV advertising?  As part of an effort to convey the effects of the growth of streaming video on the broadly-defined world of TV advertising, I’ve tried to estimate the decline in gross ratings points or time spent with advertising on TV in a world where a) streaming continues to take share of TV consumption b) consumers of the many SVOD services generally want to choose ad-free options c) most of the rest of the services will inevitably be ad-light given the lack of desirability of advertising in the on-demand world that accounts for most of streaming video.

With conservative assumptions, US television should lose 24% of inventory over the next four years.  With assumptions outlined below, I estimate a reduction in available TV inventory in the United States including linear and connected TV of around 24% between 2023 and 2027, a -6.6% CAGR.  All else held equal, this would have a positive effect on pricing that sellers of advertising experience, although because I expect spending to be down on a year over year basis every year going forward, average CPM increases might only rise more slightly, at least if averaged across all TV advertising formats and dayparts.  Note that this calculation has been made on a person 2+ basis.

Key assumptions: To arrive at this conclusion I used data from Nielsen (for historical data on time spent with different TV services), Antenna (for historical data on the percentage of subscribers choosing ad-free or ad-supported streaming services), my own historical estimates of ad loads and additional simplistic, if plausible, critical assumptions including the following:

• Total hours of television consumption will be flat between now and 2027 including YouTube and “Other” (video games or other alternative activities) uses of TV sets

• Streaming services will continue to grow viewing share over the next two years at a pace that roughly matches the increases I expect to see in the share of content spending directed to streaming over that time; share gains continue at the same pace for the next two years, adding roughly 500bps of viewing share each year.

• Amazon Prime Video excluding FreeVee and Amazon Channels covers half of all Amazon Prime Video consumption included in The Gauge.  Amazon will launch Prime Video Ads in early 2024 immediately serving ads to 100% of its Prime Video subscribers, but the share of ad-supported subscribers gradually falls to half of all Prime Video subscribers

• 20% of linear TV viewing occurs via DVRs with 50% of related ads skipped

• Other linear TV (including vMVPDs) has 20% ad loads (12 minutes per hour)

• Ad loads for ad-supported subscription tiers are assumed to be around 4 minutes per hour on Disney+, Netflix, Max, Paramount+ and Prime, 5 minutes on Tubi, 8 minutes on Hulu and Roku and 12 minutes on Pluto

• Ad-supported subscribers will amount to around approximately 30% of Netflix, Max and Paramount+ subscribers in 2027 and 50% of Disney+ subscribers.  I assume that Hulu and Peacock, both of which have been relatively committed to advertising models from the start sustain their existing ad-supported subscriber bases, with much higher penetration rates

Among the more critical assumptions are

• Changes in consumption of television.  It’s plausible that as more content is consumed on demand via streaming services, less “ambient” linear content consumption occurs.  Lower pay TV penetration could also cause this to occur

• A faster gain in share of streaming consumption.  If media companies invest more in their streaming services relative to their linear ones (or, alternately, starve their linear services more in order to support stable spending on streaming while reducing overall costs), streaming’s share of consumption would go up and ad consumption would go down

• Netflix’ advertising tier.  If more than a minority chose to watch a significant volume of content on the largest streaming service, this could conceivably change the trajectory for available ad inventory.

If the above played out, the following would be true:

• Excluding “Other” and YouTube viewing, by 2027, 54% of TV viewing would occur on streaming services with only 46% consumed on linear services (vs. 32% and 68% in 2023, respectively), including vMVPDs and cable operator streaming apps.

• Advertising’s share of TV consumption (streaming and linear TV combined) would fall from 13.1% of time spent with TV in 2023 to 10.6% by 2027

• The volume of ad-supported TV time would fall by 24% between 2023 and 2027.  This equates to a -6.6% CAGR

• If YouTube – 19% of total linear plus streaming TV consumption in 2027 at its current trajectory – were included as “TV” here, with ad loads averaging 6 minutes an hour, total ad-supported TV ad inventory under this broader definition would still be down by 13% between 2022 and 2027

Low ad loads are likely a permanent feature of streaming, and low penetration rates of ad tiers will be common on many services.  The critical issue for advertisers to be mindful of is that streaming ad loads account for only around 3% of total time spent with streaming TV, rising to more like 4% with the inclusion of Amazon Prime Video Ads and the aforementioned growth in ad-supported tiers offered by SVOD services.  By contrast, linear TV’s ad inventory accounts for around 20% of time spent with those media owners.   No matter how much streaming grows, it can never make up for lost linear ad inventory so long as ad loads remain light and consumers exhibit preferences for ad-free options for the biggest service (ie Netflix).   I believe these preferences will hold largely because in an on-demand world, the choice to consume content is highly “intentional” and thus advertising is much more of an interruption and less tolerated.   A streaming service can try to increase its ad loads, but my guess is that a service doing so by much would be likely to lose subscribers.  By contrast, consumption of linear TV features a significant volume of programming viewed in an ambient manner (i.e. as background noise or in aid of some kind of activity where multitasking is going on) and where much of sports content is consumed (and where ads fit more naturally into the flow of the programming).

On the other hand, linear TV ad inventory will remain the primary source of TV ad inventory as it’s modeled here with more than three quarters of the medium’s total available inventory by 2027.

Source: Madison and Wall

Different trends will play out for younger audiences, but marketers may increasingly focus on advanced audiences rather than demographic groups.  Importantly, the aforementioned analysis is entirely focused on all people aged two and up.    In Nielsen’s latest release of the Gauge, demographic breakdowns where provided.   Audiences aged 18-49 consume substantially more streaming content (55% of time with TVs vs. 38% for the total population), and so it’s not hard to imagine that over the course of the next five years, perhaps these younger audiences could consume 70% or more of their TV content on streaming services.   However, 18-49s account for “only” 32% of TV consumption.  Presumably their share of ad-supported content consumed will fall much more significantly vs. the overall population over the next several years to come.    As this occurs, it’s also plausible that marketers will increase their focus on so-called “advanced audiences” rather than traditional age and gender-based targets.