TV after TV: How advertisers can rebuild reach in a fragmented streaming world
Linear television is shrinking fast as audiences migrate to CTV, AVOD, FAST and YouTube on the big screen. Marketers now face rising costs, fractured measurement, and frequency waste.
Television advertising is no longer a single market. It’s a mosaic linear, BVOD, AVOD, FAST, retail and social video competing for reach, attention and measurable outcomes. Global linear spend is shrinking as viewing shifts to streaming; CTV and online video carry the growth and the headaches. The practical question for senior marketers is blunt: how do you buy reach and prove impact when “TV” is many things at once?
From monolith to mosaic
A decade ago the TV plan was a blunt instrument with two superpowers: mass reach and cultural presence. Those strengths still matter, but they live inside a more fragmented system.
Spend: Linear’s global share has fallen markedly. The “video” pie (linear plus CTV) remains sizable, but the faster compounding engines are search, social and retail media.
Audiences: Streaming is now the default growth path. Younger cohorts drift first; older ones follow more slowly. Big-screen YouTube viewing continues to climb, a reminder that “TV” is as much about the device as the channel.
Costs: Linear CPMs have inflated in some markets (notably the US) and softened in others (Japan, Brazil). Dusty global averages conceal more than they reveal.
The conclusion is not that “TV is over.” It’s that TV’s strengths must be rebuilt across multiple sellers, devices and currencies.
What counts as “TV” now?
Is TV a device (the living-room screen), a publisher class (broadcasters/streamers), or just “video” with a minimum quality bar? The industry hasn’t settled on a single definition, which creates planning frictions:
Taxonomy mismatch: Many accounting frameworks park YouTube under “online video,” not “TV,” even when viewed on a television set. That split is sensible for budget governance but confusing when reconciling spend with viewing.
Currency chaos: One platform sells watch time, another sells streams, a third sells impressions. None are directly equivalent. Without agreed conversions, econometrics flattens differences and in-flight optimisers chase the cheapest unit, not the most effective exposure.
Partial measurement: Panel, ACR and platform logs each tell a sliver of the story. Stitching them together is a craft, not a download.
Pragmatic brands solve this with a one-pager that fixes a working definition for the next 12 months and sets the reporting rules up front.
Who’s winning the “TV” dollar?
A better question: who is best placed to convert TV-like attention into provable outcomes?
Retail + video: Amazon (Prime Video) and Walmart (Vizio) hard-wire retail signals to big-screen inventory. That shortens the path from exposure to sale and speaks to CFOs in a language they trust. Expect more broadcaster–retailer clean-room tie-ups that show uplift by SKU, not just reach curves.
Device gatekeepers: Smart-TV makers own the home screen. Viewers can spend long minutes in the OS before they play anything—valuable real estate for discovery units, sponsorships and shoppable moments. This is new-to-market inventory with TV-like attention but closer to retail media in mindset.
YouTube on TV sets: On large screens, YouTube acts like a mega-network with a wider content mix—sports, creators, films—and a direct feed into Google’s performance stack. For broadcasters, YouTube is no longer just promotional plumbing; it’s a distribution and monetisation line.
Ad-tier streamers: Slower subscription growth has pushed services to ad-funded tiers. In the US they defend retention; in the UK they add incremental reach. Live events and bundles reduce churn and create more inventory with “TV feel,” but platform fragmentation persists.
Geography still matters
The US leads in streaming adoption and CTV reallocation; Europe is patchier, shaped by public-service systems, privacy regimes and later AVOD rollouts. APAC and Latin America each contain multiple realities. A global “CTV share” slide is a board comforter, not a plan. Buy country by country:
In Britain, BVOD remains prime-time heavy; daypart discipline still helps.
In Germany/France, CTV’s share grows more slowly; broadcaster alliances carry more weight.
In the US, treat CTV as the base and linear as an extension for live events and marginal reach—then measure with incrementally, not clicks.
5) The headaches that block performance
Three problems appear in every review.
Frequency waste. Within a single broadcaster, caps work; across estates, they don’t, at least not cleanly. The result is over-saturation of heavy TV viewers and under-delivery to light and streaming-first audiences. Engineering frequency across publishers is the biggest lever you haven’t fully pulled.
Currencies and definitions. If your KPI is “completions,” you’ll buy to maximise completions. If it’s “watch time,” you’ll skew to longer videos. Change the unit, change the plan. Agree definitions at briefing, not at the post-mortem.
Inflation (sometimes). US linear may be expensive; Japan and Brazil may not be. Treat CPM as local. Buy into the markets where linear is still efficient and shift the split elsewhere.
A 2025 playbook that actually ships
1) Define your “TV” and lock the rules.
Write a one-pager: Scope (BVOD, AVOD, FAST, device-home-screen units, CTV publisher inventory on TV sets), Standards (brand safety, content quality, view-ability), Accounting (where YouTube sits), Measurement (the hierarchy by decision: MMM → geo/audience lift → log-level diagnostics). Publish it internally. Stick to it.
2) Plan for reach, buy for outcomes.
Weeks 1–4: build unduplicated reach across BVOD and premium AVOD, with linear as an event-led extender. Weeks 5–12: layer retail signals (on-site and off-site) to prove incrementally. The insight isn’t new; the discipline is.
3) Make CTV the cornerstone; treat linear as an extender and an event stage.
CTV covers the base; linear covers the gaps and the tentpoles. If you still start with linear and “top up” with CTV, you’ll hammer heavy viewers and miss the rest.
4) Engineer frequency, don’t assume it.
Set a cap (for example, six to eight exposures per fortnight per household on brand flights). Enforce at IO level across estates. Use clean-room overlaps where possible; where not, model dedupe with panel + ACR + logs. The aim isn’t perfection it’s knocking out the heavy-tail waste.
5) Prove business impact with retailer truth sets.
Keep MMM for budget setting. Run geo or audience lift for campaigns that blend CTV with retailer audiences. If a channel wants credit, it must clear an incrementally hurdle. Ban last-touch ROAS as the sole proof outside pure DR.
6) Creative for the living room.
Use 15s/30s for memory building, but design for the OS and discovery moments too pause screens, home-screen placements, shoppable overlays, QR-driven offers. Test whether these formats add incremental reach or conversion at tolerable cost per lift, not whether they look cool in the deck.
The next 24 months (what will actually change)
Retail-video becomes standard.
Broadcasters and streamers will lean on retailer data to prove lift. Retailers will court brand budgets with big-screen packages tied to outcomes. Expect more Boots/ITVX-style partnerships and more off-site options linked to SKU-level reporting.
Home screens become prime inventory.
Smart-TV OS surfaces will be packaged as high-impact discovery units: programme carousels, search hubs, pause placements. This is television’s front door becoming an ad market of its own.
Frequency control sells.
The first seller to credibly de-duplicate across estates will win share. Buyers will pay a premium for time, sanity and fewer annoyed viewers.
A working framework for the C-suite
When the board asks why you changed the plan, use this three-column story.
Reach
Goal: unduplicated weekly reach at an acceptable CPM. Method: BVOD + AVOD + CTV publisher inventory, with selective YouTube TV-set buys for younger/light TV audiences. Governance: prove dedupe assumptions quarterly in your biggest market via clean-room overlaps.
Outcomes
Goal: incremental profit, not just ROAS. Method: MMM for the annual envelope; geo/audience lift for campaign proof; log-level completion/attention only where tied to sales. Governance: no channel scales on last-touch ROAS alone.
Cost control
Goal: efficiency without reach collapse. Method: market-specific CPM guardrails; flip your linear/CTV mix to fit local inflation and viewing; keep event buys, but fill everything else programmatically. Governance: optimise to market reality, not global averages.
Questions the CFO will ask (and what to say)
“Is YouTube ‘TV’?”
On the big screen it behaves like one. Keep its spend under “online video” for budget discipline, but treat it as a living-room reach tool for 16–34s and light-TV households.
“Are we over-paying for linear?”
Maybe in the US; not necessarily in Japan or Brazil. Buy linear where it’s still efficient and where live events add cultural weight. Elsewhere, CTV should be the base.
“Will CTV budgets rise again?”
Intent is still up particularly in the Americas because usage outruns spend. Europe and APAC grow more cautiously. Your plan should reflect that patchwork.
“Do ad tiers just cannibalise subscriptions?”
They tend to improve retention in the US and add incremental reach in the UK. The winners use live content and bundles to keep both tiers sticky.
A plain-spoken close
TV’s core promise hasn’t changed: quality reach that moves real-world numbers. What’s changed is where that reach lives and how you prove it. The winners will treat CTV as the base, linear as the event-led extender; they’ll enforce frequency, not merely report it; they’ll prove outcomes with retailer truth sets instead of leaning on last-touch alibis; and they’ll design creative for the living-room experience, including the home screen. The mosaic is the market now. Build your plan to match it.
Thanks for reading all the way to the end of the article! This post is public so feel free to share it, and if you have not done so already sign up and become a member.1
The article is based on the WARC report Global TV Ad Trends available on the 9th September 2025