Connected television has become one of the most important forces reshaping media planning because it changes both how television is delivered and how television inventory is bought, targeted, and measured. In the simplest sense, connected television, or CTV, refers to television content accessed through internet-connected devices such as smart TVs, streaming devices, gaming consoles, and connected set-top boxes. That definition matters because CTV is not a genre of content and it is not synonymous with streaming in every context. It is a delivery environment in which viewers watch long-form video through apps and services rather than through a traditional cable or broadcast signal alone.

For media planners, that shift creates opportunity and complexity at the same time. CTV can offer the sight, sound, and screen presence long associated with television, but it operates through a more fragmented distribution system with different buying mechanisms, identity assumptions, measurement practices, and frequency risks. As television audiences disperse across broadcast networks, cable channels, subscription streaming services, free ad-supported streaming television services, virtual MVPDs, and ad-supported apps, planning television is no longer mainly a matter of selecting networks and dayparts. It increasingly requires understanding how inventory is distributed across platforms, what counts as an impression, how household targeting works, where duplication occurs, and how CTV complements or competes with linear television in pursuit of reach.

A useful starting point is to distinguish CTV from linear television and from streaming as a broader consumer behavior. Linear television is scheduled programming delivered in a fixed stream, whether through broadcast, cable, or a digital multichannel distributor. Streaming refers to video delivered over the internet, whether live or on demand, ad-supported or subscription-based. CTV refers specifically to viewing that happens on an internet-connected television device. Someone can stream on a phone or laptop without that being CTV. Conversely, some CTV inventory is a live stream of linear channels. These distinctions matter because advertisers are not just choosing content. They are choosing a media environment with different audience signals, ad loads, transaction structures, and measurement rules.

Audience behavior is the reason CTV now sits at the center of television planning. According to Nielsen’s monthly The Gauge reports, streaming has accounted for a larger share of total U.S. television usage than broadcast or cable for extended periods, reflecting the continuing migration of viewing time toward app-based television environments. Nielsen’s methodology documentation makes clear that these are usage shares of TV time, not advertising exposure shares, but the trend is strategically significant because time spent affects where inventory supply accumulates and where advertisers must look for incremental reach. See https://www.nielsen.com/insights/2024/the-gauge/.

That shift does not mean linear television has ceased to matter. Broadcast and cable still provide scale, cultural concentration, sports rights, local news, and predictable mass audiences that many streaming environments cannot replicate consistently. Major live events remain especially important because they aggregate audiences in real time and often command premium pricing. For planners, the practical question is not whether CTV replaces linear TV. It is how each contributes differently to reach, frequency, audience composition, and campaign timing.

CTV changes media planning first by changing the unit of audience access. Traditional television planning has long relied on program schedules, network ratings, daypart patterns, and demographic guarantees. CTV planning begins more often with an inventory graph made up of streaming publishers, platform operators, device ecosystems, and ad-tech pathways. Some inventory is sold directly by premium publishers such as network-owned streaming services. Some is sold through demand-side platforms in private marketplaces or programmatic guaranteed arrangements. Some comes through device makers or operating systems that aggregate inventory or data signals. Some FAST inventory behaves more like a digital video marketplace, with large volumes of lower-cost avails spread across numerous channels and apps. The result is that two CTV impressions may look similar on the living room screen while being commercially and operationally very different.

This fragmentation complicates planning because content availability, audience composition, and measurement transparency vary widely by seller. A high-quality long-form episode on a major broadcaster’s ad-supported app is not interchangeable with a lightly curated FAST channel stream, even though both may be counted as CTV impressions. Ad loads differ. Completion rates differ. Context differs. Viewer expectations differ. So do pricing and controls. Treating CTV as one undifferentiated pool of “streaming TV” can produce misleading assumptions about value and audience quality.

Targeting is one of the most discussed advantages of CTV, but it is often misunderstood. Compared with traditional age-sex television buying, CTV can offer more flexible audience selection using household-level data, geographic filters, behavioral segments, first-party data matches, and, in some cases, retail or purchase-based audiences. Yet CTV targeting is not the same as one-to-one individual targeting in the way many digital display buyers imagine. In most cases, CTV advertising is delivered to a household or device environment, not to a verified individual person. If one household streams through a smart TV app, the ad opportunity may be associated with a household ID, IP-based signal, authenticated account, or device graph, but that does not guarantee knowledge of who in the room actually saw the ad.

That distinction matters strategically. Household identity can be valuable when the product is purchased jointly, when broad household attributes are relevant, or when advertisers want to connect living room exposure to other devices in the home. But it also introduces ambiguity, especially for products with narrow individual targets or significant age and gender skews within the same home. A planner buying against “households with children” or “auto intenders” is making a different kind of media decision from one buying adults 25 to 54 on linear ratings. Neither approach is inherently superior. Each rests on different assumptions about who is actually exposed.

Frequency management is where the promise of CTV often runs into operational reality. Because CTV inventory can be bought across many publishers, apps, device makers, and exchanges, the same household may receive repeated exposure from different supply sources that do not share complete identity information. Even when a platform can cap frequency within its own environment, that control may not extend across the full campaign. A household exposed on Hulu, Roku, Peacock, YouTube TV, and several FAST apps may experience much higher actual frequency than campaign averages suggest. This is especially likely when planners layer multiple demand paths to gain scale.

That problem is not unique to CTV, but it is intensified by fragmentation and inconsistent identity resolution. Average frequency at the campaign level can conceal a highly uneven distribution in which some households are lightly exposed while others are hit repeatedly. In media planning terms, that can reduce incremental reach and create waste, particularly when CTV is bought as if every platform delivers a distinct audience. Frequency management therefore becomes less about selecting an ideal number and more about controlling exposure distribution across a fragmented supply chain.

Cross-platform planning makes the issue even more complex. A campaign may run on linear TV, broadcaster streaming apps, digital video platforms, and social video at the same time. The planner’s challenge is not simply to total impressions but to estimate deduplicated reach and understand whether CTV is extending the television audience or merely repeating exposures against viewers already reached elsewhere. This is one reason cross-media measurement has become so central to the CTV conversation.

Measurement in CTV remains powerful but imperfect. Traditional television has long depended on panel-based audience measurement, with ratings used as a currency for many transactions. CTV introduces new data sources, including ad-server logs, automatic content recognition from smart TVs, return-path data from set-top boxes, device graphs, and publisher first-party data. These sources can expand visibility into exposure patterns, but they do not eliminate uncertainty. Logs can show that an ad was served to a device or app environment. They do not prove that a specific person watched attentively. Automatic content recognition can detect content exposure from participating devices, but not from every screen or every platform. Identity graphs connect devices and households probabilistically or deterministically depending on inputs, which means deduplication often involves modeling rather than direct observation.

The Media Rating Council’s accreditation process and methodological standards are relevant here because they underscore that data quality depends on defined methods, coverage, calibration, and auditability rather than on claims of “census-level” scale alone. Large device datasets can still have representation gaps and require weighting to estimate people-based audiences. More data does not automatically mean better audience truth if the observed population is biased or if identity stitching introduces error. See the Media Rating Council overview at https://mediaratingcouncil.org/.

Impressions in CTV also need careful interpretation. In ad-supported streaming, an impression typically represents the serving of an ad according to the delivery rules of that platform or measurement system. It does not mean a unique viewer, guaranteed attention, message recall, or business impact. In many premium CTV environments, completion rates are high because ads usually play full screen in long-form content and are harder to skip than many desktop or mobile video ads. Even so, a completed ad is not the same thing as active attention. Viewers may leave the room, use a second screen, mute the set, or disengage mentally during commercial breaks. CTV generally offers stronger opportunity-to-see conditions than much digital display inventory, but opportunity should not be confused with proof of cognitive attention.

Viewability, a standard concept in digital media, is often less central in CTV discussions because television-format video is usually served in a full-screen player on a television set. Yet the larger issue remains the same: technical delivery does not equal human attention. Advertisers interested in attention metrics should examine how providers define them, whether they rely on screen position, completion, audio state, co-viewing assumptions, or modeled probabilities, and whether those measures are comparable across environments. Attention can be a useful planning lens, but it is not a universal currency.

Buying methods in CTV reflect the hybrid nature of the market. Some inventory is bought much like television, through direct negotiations, annual commitments, and premium publisher relationships. This is common for major ad-supported streaming services, live sports, and high-demand premium video environments. Some is bought through programmatic pipes using private marketplaces, preferred deals, or programmatic guaranteed arrangements. Some is transacted in open auctions, particularly within broader streaming and FAST supply. Each method carries different implications for price, quality control, transparency, and delivery predictability.

Programmatic CTV is often presented as inherently more precise or efficient, but that oversimplifies the market. Automation can improve workflow, enable faster optimization, and give buyers access to audience data overlays or supply-path controls. It can also introduce hidden fees, variable transparency, inconsistent content classification, and uneven inventory quality. A premium direct deal may provide stronger content assurance, clearer makegood procedures, and better access to sponsorship integrations. A private marketplace may balance control with efficiency. An open auction may be useful for incremental scale but can raise more questions about duplication and supply-chain transparency. The correct buying method depends on the objective, budget, need for contextual control, and tolerance for measurement uncertainty.

Economics matter here because CTV is often evaluated through CPM comparisons that flatten meaningful differences. Premium CTV inventory commonly commands higher CPMs than many forms of digital video because advertisers are paying for a scarce combination of television-style environment, larger screen experience, and increasingly valuable audience reach among light linear viewers. But a lower CPM elsewhere does not automatically represent better value if it delivers weaker attention conditions, poorer context, or heavily duplicated audiences. Conversely, a premium CTV CPM is not justified simply because the ad appears on a television screen. The strategic question is what audience is being reached, under what conditions, with what level of duplication, and at what incremental contribution to campaign goals.

CTV also changes geographic planning. Traditional local television remains a strong vehicle for market-by-market delivery, especially for retail, healthcare, political advertising, and regional services. CTV can support geographic targeting at national, regional, or local levels, sometimes with more flexibility than linear schedules. However, local CTV supply is not always as robust or standardized as local broadcast inventory, and local delivery can be constrained by app scale, publisher footprint, or household matching quality. Advertisers should not assume that CTV automatically replicates all the strengths of local TV.

Publisher and platform economics further shape the planning landscape. Ad-supported streaming has become strategically important because content companies need multiple revenue streams as subscription growth matures and carriage economics evolve. Some services operate entirely as advertising-supported businesses. Others mix subscription revenue with ad tiers. FAST services rely heavily on ad monetization and often expand inventory supply with lower barriers to consumer entry. Device makers and operating systems also play a larger role than they did in traditional television because they can control home-screen interfaces, aggregate inventory, collect usage data, or influence ad technology integration. This redistributes power across the television value chain. Media planners therefore increasingly negotiate not just with networks, but with streamers, platform intermediaries, operating systems, and ad-tech vendors.

This changing power structure has strategic consequences. Inventory that appears to belong to a content brand may in practice be sold or mediated through multiple channels. Measurement and frequency control may depend on cooperation among parties with different incentives. Publishers seek yield and audience monetization. Platforms seek data advantages and transaction share. Buyers seek deduplicated reach and transparent delivery. The planner’s task is to understand where control actually resides in the supply chain.

The relationship between CTV and linear TV is best understood as complementary but uneven. Linear still excels at broad, fast reach, especially around live programming, sports, tentpole events, and older audience segments with stable viewing habits. CTV is often used to extend that reach among light linear viewers, younger streaming-heavy households, and audiences that are difficult to assemble efficiently through schedule-based television alone. In some campaigns, linear provides scale and cultural concentration while CTV adds precision and incremental reach. In others, especially for brands targeting cord-cutters or digitally fluent audiences, CTV may carry a larger role and linear may become selective support rather than the core buy.

But that complementarity should not be assumed. There are cases where CTV simply reproduces television’s audience without adding much new reach, especially when the same media owners distribute content across both linear and streaming outlets and when cross-screen duplication is high. There are also cases where linear still delivers better cost efficiency for mass reach because its audience aggregation remains stronger in certain programming environments. Effective planning therefore depends on disciplined analysis of audience overlap, not on channel fashion.

For planners building a television strategy today, the most useful approach is to start with the objective and work outward. If the task is rapid national awareness, live-event concentration, or broad age-based reach, linear may still do a large share of the work. If the task is to extend coverage into streaming-first households, improve audience selectivity, or use household data in a television environment, CTV becomes more important. If the task is local retail activation, planners must examine whether local broadcast, local cable, CTV geo-targeting, or some combination offers the best balance of coverage and control. The point is not to declare one form of television superior. It is to understand the distinct media mechanics behind each.

What CTV ultimately changes is not the need for television planning but the skill set required to do it well. The planner must now think simultaneously like a television buyer, a digital video strategist, and a cross-platform measurement analyst. That means understanding publisher quality, identity assumptions, supply paths, frequency controls, impression definitions, and household-level limitations, while still applying classic television disciplines such as reach building, schedule construction, audience composition, and context evaluation.

Connected television has made television more targetable, more data-rich, and in some respects more accountable. It has also made television more fragmented, less standardized, and harder to control cleanly across platforms. For advertisers, that is the central planning reality. CTV expands the television opportunity, but it does not simplify it. Better decisions come from treating CTV not as “digital plus a big screen” and not as “TV with better targeting,” but as a hybrid media environment whose value depends on how carefully its audience access, inventory sources, buying mechanics, and measurement limits are understood.