CTV Programmatic: The $43B Opportunity Most Buyers Are Still Getting Wrong

Connected TV programmatic advertising is projected to exceed $43 billion in US spend by the end of 2025, making it the fastest-growing channel in the programmatic ecosystem. Despite that scale, a significant majority of CTV campaigns continue to underperform, not because the channel doesn't work, but because buyers are applying display-era strategies to a fundamentally different medium with fundamentally different economics, audience behaviors, and measurement mechanics.

After analyzing campaign performance data across thousands of CTV campaigns running through the Adbite exchange, the pattern is unmistakable: the buyers who get CTV right approach it as a completely distinct medium. The buyers who struggle are those who treat it as "video on a big screen."

The CTV Market: What the Numbers Actually Say

The $43 billion headline obscures as much as it reveals. CTV ad inventory is highly fragmented across streaming platforms, smart TV operating systems, and connected device manufacturers, each with different ad formats, targeting capabilities, measurement systems, and auction dynamics. Roku, Amazon Fire TV, Samsung Tizen, and LG webOS each operate their own first-party data ecosystems. AVOD platforms like Peacock, Paramount+, and Max have separate programmatic access points with varying floor prices, frequency caps, and creative specifications.

The programmatic portion of CTV, impressions transacted through open auction or programmatic guaranteed deals rather than direct IO, represents approximately 60% of total CTV spend and growing. The open auction share is more complex to buy well, but offers the scale and targeting flexibility that makes programmatic CTV compelling for performance-oriented advertisers and brand campaign managers alike.

Why Display Buying Strategies Fail in CTV

The most common failure mode in programmatic CTV is the application of audience retargeting logic that was designed for display inventory. In display, retargeting works because users are actively browsing, they can click immediately on a served ad, have high purchase intent, and are accustomed to direct response creative. The optimization signal is fast: click, conversion, attributed revenue. The feedback loop is short.

CTV is a fundamentally passive medium. Users are watching content they've chosen to engage with for extended periods, streaming dramas, sports, news programming. They are not browsing. They cannot click. Even when CTV ad units include QR codes or companion mobile experiences, the direct interaction rate is a fraction of display. Retargeting audiences in CTV creates frequency fatigue without a proportional conversion uplift, burns budget on users who were already highly likely to convert through organic channels, and generates poor view completion rates when creative isn't adapted to the medium.

The second failure mode is bidding strategy. Display programmatic buyers have learned to bid precisely, using frequency caps, audience exclusions, and domain-level targeting to hit specific CPM targets with high fill efficiency. CTV inventory is scarcer, less liquid, and priced at significantly higher CPMs (often $25-60+ for premium streaming content versus $1-5 for display). Display-calibrated bidding logic, tight CPM caps, aggressive floor rejection, results in systematic underdelivery on CTV, because the inventory that clears at display prices isn't the inventory that drives brand impact.

Third: creative strategy. The 30-second unskippable pre-roll that dominates CTV is a completely different canvas from the 300x250 display unit. Creative that doesn't engage the viewer within the first five seconds, before the skip window in skippable formats, will see completion rates collapse. Most display buyers bring repurposed digital video creative into CTV without adaptation, generating completion rates 30-40% below benchmarks for the channel.

The Audience-First Approach to CTV Buying

High-performing CTV buyers start with audience definition rather than inventory selection. The question isn't "which streaming services should we run on?", it's "who are we trying to reach, and where do we have the highest probability of reaching them with relevance?" For a CPG brand targeting mothers with children under 12, the answer might be family content across multiple AVOD platforms identified through ACR data signals. For an automotive brand targeting in-market car shoppers, the answer might be sports content where audience data shows higher correlation with recent dealership visits.

Automatic Content Recognition (ACR) data is the most powerful targeting signal available in CTV. ACR technology, deployed in smart TVs by manufacturers including Samsung, LG, and Vizio, identifies what content a user is watching at the household level by matching audio and video fingerprints against a content database. This generates granular viewing data, not just "watches Netflix" but "watched the first three episodes of a new drama series this week", that enables precision audience targeting based on actual content consumption.

ACR-based audiences can be used to reach households that watch specific genres, shows, networks, or competitive advertising. A car brand can target households that have recently been exposed to competitor automotive ads, a powerful conquesting application. A pharma brand can reach households where linear TV viewing patterns suggest the presence of an older adult demographic without relying on cookie-based targeting that doesn't exist in CTV.

Frequency Capping: The CTV-Specific Challenge

Frequency management in CTV is one of the channel's most technically complex challenges. In display programmatic, frequency capping is straightforward: most DSPs implement user-level frequency caps via cookie or device ID. In CTV, the household is the relevant unit, multiple people share a single screen and a single device ID, and the fragmented nature of the ecosystem means that frequency caps set in one DSP don't communicate with frequency caps in another DSP running concurrent campaigns.

The result is a common and deeply frustrating experience for CTV viewers: being served the same 30-second ad six times in a row across three different streaming sessions. From the advertiser's perspective, this represents significant wasted spend, you're paying for the fourth through sixth impression when the first three already generated all the brand recall the impression was capable of producing. From the audience perspective, repetitive ad exposure is one of the primary drivers of streaming subscription preference over ad-supported tiers.

Best-practice CTV buyers manage frequency at the campaign level across all DSPs, using a unified identity layer to deduplicate household exposures across platforms. This requires working with SSPs and exchanges that support cross-publisher frequency management, a capability that is available on Adbite's CTV exchange but remains uncommon in the broader ecosystem. Until cross-ecosystem frequency management is standardized, sophisticated buyers use pacing controls, publisher-level frequency caps, and deal-level constraints to approximate household-level frequency management.

Measurement: The ACR Advantage and Co-Viewing Reality

CTV measurement is categorically different from display or mobile. There is no click-through rate. Viewability as defined for display (50% of pixels in view for 2 seconds) is meaningless for a full-screen TV experience. The relevant metrics are: video completion rate (VCR), specifically 25%, 50%, 75%, and 100% quartile completion; brand lift measured through survey methodology; foot traffic lift for retail advertisers; and downstream digital activity (web visits, search queries) attributable to CTV exposure through IP-to-household matching.

ACR data is transformative for CTV attribution specifically because it enables closed-loop measurement at the household level. A retailer can match households exposed to a CTV campaign, identified via ACR, against households that subsequently made an in-store or e-commerce purchase. The match rates are significantly higher than cookie-based attribution because ACR operates at the household level with persistent identifiers not affected by browser privacy settings or cookie deletion.

Co-viewing, the reality that multiple people watch the same CTV screen simultaneously, is routinely ignored in CTV measurement and is one of the most significant sources of measurement error. A family of four watching primetime programming together represents four ad exposures per 30-second spot, but most impression-level measurement counts one. Nielsen's co-viewing index data suggests that premium content in primetime has an average co-viewing multiplier of 2.3x. Buyers who fail to account for co-viewing are systematically undervaluing their CTV reach.

How to Get CTV Buying Right

The practical framework for high-performing programmatic CTV comes down to five principles. Start with audience, not inventory, define who you're reaching and use ACR and first-party data to find them. Budget for CTV CPMs, don't apply display floor logic to a channel that clears at 10-20x display CPMs. Adapt creative for the medium, invest in CTV-specific 15 and 30-second creative optimized for lean-back attention, with clear brand recall within the first five seconds. Manage frequency aggressively, use every available cross-publisher frequency tool and accept that some impression loss is worth avoiding overexposure. Measure what matters, completion rate, brand lift, and downstream attribution; not CTR.

CTV programmatic is genuinely a $43 billion opportunity. The buyers who capture disproportionate value from it will be those who develop CTV-native buying expertise rather than importing their display playbook. The medium rewards those who learn its specific dynamics, and consistently punishes those who don't.