BT FY 2005-06 Q4 results

We forecast broadcaster viewing to decline to 50% of all video viewing by 2031 (from 54% in 2025), driven by continued declines in live viewing, albeit at a slower rate than the sharp year-on-year drop seen in 2025.

Non-live broadcaster viewing will increase its share, supported by older demographics as viewing plateaus among under-35s—a trend also true of SVOD services. YouTube is projected to continue its advance on the TV set across all ages.

Although total TV ‘usage’ is flat versus pre-COVID levels, a material and seemingly growing proportion of it (~9%) is when nothing is being watched (e.g. pausing, browsing): this is correlated with smartphone usage and impinges on actual viewing volumes.

The middle tier is gone. Streaming pricing is bifurcating toward premium ad-free and ad-scaled tiers. Netflix played the long game—letting Disney+ and HBO Max normalise higher prices and absorb the churn risk, then hiking Standard to $19.99 and Premium to $26.99 to reclaim the premium position once the ceiling had been raised.

The industry has crossed the profitability threshold, but the gap with Netflix widens rather than closes. Linear is still funding the transition—the spin-offs and segment collapses are a visibility play as much as a structural one. 

Ad tiers are scaling fast but monetising slowly. Sports is the acquisition flywheel everyone is betting on, yet the ROI remains unproven—Peacock's losses widened the moment NBA rights hit. Reach is rising; depth of engagement is diverging. 

OpenAI has switched off its video AI Sora, under pressure to streamline and monetise as Anthropic captures enterprise momentum.

Google remains as the dominant US video AI player, with more established media industry relationships. Video AI will coalesce around more targeted, industry-ready tech.

The media industry might not mourn Sora, but the unwinding of the Disney/OpenAI deal means it's back to the drawing board on establishing a model for extracting revenue when media IP is used in video AI.