Video advertising has undergone a fundamental structural transformation. What was once a medium defined by scheduled broadcast slots and fixed audience demographics has become an algorithmically orchestrated, cross-screen experience that reaches consumers at moments of genuine intent. The numbers reflect this shift with startling clarity. According to eMarketer, global digital video advertising expenditure surpassed $191 billion in 2024 and is projected to exceed $290 billion by 2028, a compound annual growth rate that comfortably outpaces every other digital channel.
Driving this expansion are three converging forces: the maturation of connected television (CTV) as a premium inventory environment, YouTube’s sustained dominance as both a search engine and entertainment platform, and the proliferation of programmatic infrastructure that has made real-time video buying as fluid as display. Understanding how these forces interact — and where the technological architecture underpinning them is headed — is essential for any organisation seeking to deploy video budgets with precision.
The YouTube Ecosystem: Scale Meets Signal
YouTube remains the single largest video advertising platform on earth, reaching more than 2.7 billion logged-in users monthly according to Alphabet’s 2024 earnings disclosures. What distinguishes YouTube from broadcast and even most streaming competitors is not reach alone but the quality of its intent signals. Users arrive at YouTube with explicit curiosity — searching for product reviews, tutorials, comparisons — which means the platform’s targeting infrastructure can match advertising messages to audiences at moments of genuine decision-making rather than passive viewing.
Alphabet’s Performance Max campaigns, which deploy machine learning to allocate budget across YouTube, Search, Display, Discover, and Gmail simultaneously, generated significantly above-average return on advertising spend for retail advertisers in Alphabet’s own published case studies during 2024. The mechanism is straightforward: Google’s AI continuously appraises conversion probability for each impression and adjusts bid values in real time, optimising across formats including skippable in-stream, non-skippable bumper ads, and the increasingly prominent YouTube Shorts inventory.
YouTube Shorts represents a strategic counter to TikTok’s dominance in short-form video. Alphabet reported 70 billion daily Shorts views in early 2024, and advertising monetisation of that inventory — through six-second non-skippable ads and Shorts feed placements — is accelerating. For brands targeting Generation Z and younger Millennials, Shorts provides reach at CPMs that are currently more competitive than TikTok due to lower demand-side saturation.
Connected Television: The Premium Frontier
Connected television is the most consequential structural shift in video advertising since the advent of real-time bidding. CTV refers to television sets equipped with internet connectivity — smart TVs, streaming sticks, gaming consoles — through which viewers access streaming services such as Netflix, Disney+, Peacock, Paramount+, and a proliferating ecosystem of free, ad-supported streaming television (FAST) channels.
GroupM’s 2024 global advertising forecast placed CTV advertising at approximately $35 billion, with a projected trajectory toward $60 billion by 2027. The appeal for advertisers is layered. First, CTV combines the lean-back attention environment of traditional television with the addressability of digital — an advertiser can target a household based on income bracket, purchase history, and content genre preference simultaneously. Second, because streaming services now span the full spectrum from subscription-only to fully ad-supported, inventory availability has expanded dramatically. Netflix’s ad-supported tier, launched in late 2022 and achieving 40 million monthly active users by mid-2024 according to the company’s own disclosures, is now a material part of global CTV planning.
The Trade Desk, Magnite, and FreeWheel have each invested heavily in CTV-specific supply-side infrastructure. The Trade Desk’s OpenPath initiative, which enables direct connections between demand-side platforms and premium CTV publishers, bypasses traditional intermediary layers and allows buyers to access inventory with greater transparency over exact placement and audience composition. This architecture matters enormously given CTV’s higher CPMs — frequently £20 to £50 per thousand impressions in the UK market — which make fraud and placement quality consequential financial risks.
Programmatic Video: Infrastructure and Innovation
The programmatic infrastructure supporting video has evolved considerably from its display-advertising origins. Video’s complexity — variable ad lengths, diverse player environments, multiple screen sizes, buffering and viewability challenges — demanded purpose-built protocols, and the industry has largely settled on a suite of standards promulgated by the Interactive Advertising Bureau.
VAST (Video Ad Serving Template) and VPAID (Video Player Ad Interface Definition) standardised the communication layer between ad servers and video players. More recently, OMID (Open Measurement Interface Definition) has become the measurement standard that allows third-party verification vendors such as Integral Ad Science and DoubleVerify to confirm viewability and brand safety in video environments without requiring proprietary SDKs in every player. The IAB’s OpenRTB 2.6 specification, ratified in 2022 and adopted by major exchanges through 2024, extended bid request parameters specifically for CTV and out-of-stream video, enabling more granular audience and context signals to accompany each impression offer.
From a buyer’s perspective, programmatic video operates through two primary channels: programmatic guaranteed deals, in which inventory is reserved at a fixed CPM through a direct negotiation that is then executed programmatically; and open marketplace bidding, in which publishers offer impressions through real-time auctions. For premium CTV inventory, programmatic guaranteed and private marketplace deals dominate, because publishers wish to maintain floor prices and control over brand adjacency. For mid-tier and FAST channel inventory, open marketplace trading is more prevalent.
Measurement: The Industry’s Defining Challenge
No area of video advertising technology generates more urgency — or more controversy — than measurement. The linear television industry constructed its entire commercial model on a single currency: Gross Rating Points, derived from Nielsen’s panel methodology. Digital video fragmented that currency across dozens of competing measurement vendors and platform-native attribution tools, creating conditions in which the same campaign could produce wildly different reported performance depending on who was counting.
Nielsen’s 2024 introduction of Nielsen ONE, a cross-media measurement solution intended to unify linear and digital audience counting, represents the most ambitious attempt yet to reconstruct an industry currency for a fragmented era. Simultaneously, iSpot.tv, VideoAmp, and Comscore have positioned themselves as alternative currencies, each attracting major broadcaster and advertiser commitments to pilot cross-platform reach and frequency measurement.
The technical challenge is significant. Matching linear television viewing (which Nielsen still measures primarily via household panels) with CTV streaming data (which carries device-level identifiers) and digital video data (which increasingly operates under post-cookie identity frameworks) requires probabilistic matching at scale. VideoAmp’s measurement methodology, for example, combines set-top box data from cable operators, ACR (automatic content recognition) data from smart TV manufacturers such as Samsung and LG, and deterministic identity graphs to construct a unified view of reach across screens.
Attention Metrics and Beyond
Forward-looking advertisers and platforms are increasingly supplementing traditional viewability metrics — which simply confirm whether 50 per cent of an ad was visible for two seconds — with attention metrics. Adelaide, a New York-based measurement company, developed a metric called AU (Attention Unit) that incorporates eye-tracking studies, panel data, and publisher environment quality signals to produce a single score predicting the attention each impression is likely to capture. Initial adopters including Publicis Groupe and GroupM have reported improved campaign outcomes when attention scores are layered onto programmatic buying strategies.
Eye-tracking research from Lumen Research, which has conducted studies across major UK broadcasters and streaming platforms, consistently finds that CTV advertising captures substantially higher active viewing rates than in-stream video on mobile devices, where thumb-scrolling and multitasking reduce dwell time. This attention premium is a significant part of the case for CTV’s higher CPM rates.
The Short-Form Revolution: TikTok, Reels, and Shorts
The explosive growth of short-form video — defined loosely as content under 60 seconds designed for vertical mobile consumption — has created a distinct advertising environment that operates by different psychological and algorithmic rules than long-form video.
TikTok’s advertising revenue reached an estimated $18 billion globally in 2024, according to Insider Intelligence estimates, driven by TopView placements, In-Feed ads, and its Spark Ads format, which allows brands to amplify organic creator content as paid promotion. The platform’s For You Page algorithm, which distributes content based on engagement signals rather than follower graphs, means that well-crafted branded content can achieve disproportionate organic reach when user engagement rates are high.
Meta’s Reels placements across Instagram and Facebook have become a material revenue line for the company. Meta reported that Reels monetisation efficiency reached parity with its main feed inventory on Instagram in 2024, resolving a concern that had weighed on investor sentiment during the initial Reels push. For advertisers, Reels offers the demographic reach of Instagram’s established user base with the creative format flexibility of short-form video.
Contextual and AI-Powered Creative Optimisation
Advances in AI are reshaping both creative production and contextual targeting within video advertising. Dynamic Creative Optimisation (DCO) — which assembles personalised video ad variations from modular creative components in real time — has moved from experimental to mainstream. Clinch, Flashtalking (now Mediaocean), and Adobe Advertising Cloud all offer DCO capabilities that can serve thousands of creative variations based on audience segment, time of day, weather, and device type.
On the contextual side, the deprecation of third-party cookies has accelerated investment in video-specific contextual targeting. Companies such as GumGum deploy natural language processing and computer vision to analyse video content frame-by-frame and audio track-by-track, enabling keyword-level contextual targeting without any audience data. A financial services brand, for example, can target video content discussing investment decisions without relying on third-party audience segments — a capability that is both privacy-compliant and often highly effective given the alignment between content consumed and consumer intent.
What Comes Next
Several developments will define the next phase of video advertising technology. Server-side ad insertion (SSAI), which stitches ads directly into the video stream at the server level rather than calling a separate ad tag at the player level, is becoming the industry standard for CTV delivery. SSAI eliminates buffering gaps between content and ads, dramatically reduces ad-blocking efficacy, and enables more sophisticated dynamic ad loads that can adjust based on real-time audience and inventory conditions. Brightcove, AWS Elemental, and Yospace are among the leading SSAI infrastructure providers.
Interactive video advertising — which allows viewers to engage with in-stream units via remote control or touch interface — is gaining traction, particularly in the UK market where broadcasters including Channel 4 and ITV have invested in interactive overlays for their BVOD (broadcaster video-on-demand) platforms. Shoppable CTV, in which viewers can add products to a cart or receive promotional QR codes directly from an advertisement without leaving the content, represents the most commercially ambitious application of this interactivity.
The convergence of generative AI with video production will likely compress creative timelines and enable unprecedented personalisation at scale. Tools from companies such as Runway, Sora (OpenAI), and Adobe Firefly are being evaluated by major agency holding groups for their potential to generate market-specific video variants from a single master creative, at a fraction of the traditional post-production cost.
Video advertising technology stands at an inflection point where the creative, the technical, and the commercial are more tightly integrated than at any previous moment. Organisations that invest in understanding the full stack — from SSAI infrastructure to attention measurement to generative creative — will be positioned to extract material competitive advantage as the medium continues to grow.
Previously covered in this series: Connected TV Advertising: Streaming, Smart TVs and the Future of Television Buying | AI Targeting in AdTech | Real-Time Campaign Analytics | Social Commerce Advertising
