Connected TV ads put video advertising on the living-room screen through the internet rather than a cable feed: smart TVs, streaming devices and gaming consoles playing ad-supported apps and services. For a buyer the appeal is obvious, television-scale attention bought with digital-style targeting and measurement, but the market is layered with platforms, resellers and agency mark-ups, so the practical questions are where the inventory actually comes from, what targeting and measurement genuinely exist, and how to buy without paying for mystery. This page uses Google's own advertising documentation as the reference case, because it is public and checkable, and the same questions transfer to any CTV platform an agency proposes.
What counts as connected TV inventory
In Google's advertising documentation, TV screens are a device category in their own right: devices that stream TV content such as smart TVs, gaming consoles, and connected devices like Chromecast. That definition matters to buyers because it separates connected TV from the adjacent things agencies sometimes blur into it: it is not linear broadcast bought programmatically, and it is not ordinary in-app video that happens to run on a phone. The inventory is streaming video watched on a television, and platforms treat it as its own device type with its own targeting and bidding controls. When an agency proposes a CTV plan, the first clarifying question is which apps, services and device types the spend will actually appear on, and whether the plan can report by device category so television delivery is visible rather than folded into a blended video number.
How the buying works on Google's side
Google exposes connected TV through two documented routes. Device targeting lets Display and Video campaigns target the TV screens category specifically, with bid adjustments per device, so an advertiser can weight delivery toward televisions or exclude them entirely; the documentation is explicit that this targeting option is only available for Display and Video campaigns. Separately, the Google TV network extends video campaigns to in-feed inventory on the Google TV home screen, bought through video reach campaigns with an efficient-reach subtype, with audience segment targeting based on interests and habits plus location targeting, and creative requirements of horizontal 1920 x 1080 video in 16:9. The buyer's takeaway is that CTV on Google is ordinary campaign machinery with a device lens, not a separate black box, and any agency running it should be able to show the campaign types, targeting settings and device reports directly in the account.
Measurement, honestly stated
Connected TV sits between brand television and performance digital, and honest measurement admits it. The screen is shared, so an impression is a household exposure more than a personal one, and the viewer usually cannot click; platforms bridge that gap with mechanisms like on-screen QR codes for direct response, and with reach planning tools that measure connected TV alongside mobile, desktop and tablet. What a buyer should require is stated plainly in the plan: which outcomes the campaign is accountable for (reach and frequency against a defined audience, or measurable response such as scans, visits and searches), how incrementality will be judged, and what the reporting will show per device category. An agency that sells CTV on performance language but reports only impressions has moved the goalposts between the pitch and the invoice.
Questions that protect the budget
Four questions do most of the work. Where does the inventory come from, listed by platform and app, and can delivery be reported at that level? What is the full cost stack, platform fees, data fees and the agency's own margin, stated separately from media? Who owns the ad accounts and the data, so the advertiser can change agencies without losing history? And what does the first measurable milestone look like, on what date, judged by which report? None of these are CTV-specific, which is the point: connected TV is a young channel wrapped around old buying disciplines, and the agencies worth hiring answer old-discipline questions without flinching. As with search, no vendor can guarantee an outcome; what they can guarantee is transparency of delivery, cost and measurement, and that is what the contract should bind.
Questions people ask about connected tv ads
What devices do connected TV ads run on?
Google's documentation defines the TV screens category as devices that stream TV content: smart TVs, gaming consoles and connected streaming devices such as Chromecast. Campaigns can target or exclude that category and adjust bids for it.
Can small advertisers buy connected TV ads?
Yes. On Google's side, Display and Video campaigns can target TV screens with ordinary campaign budgets, and the Google TV network is bought through standard video reach campaigns. The barrier is producing broadcast-quality 16:9 video, not a minimum spend.
How do viewers respond to a TV ad they cannot click?
Direct response is bridged with mechanisms like on-screen QR codes, which Google supports for connected TV, alongside measuring searches and visits that follow exposure. Agree upfront whether the campaign is judged on reach or on measurable response.
What should an agency report for a CTV campaign?
Delivery by device category and platform, reach and frequency against the defined audience, the full cost stack separated from media, and the response metrics agreed at kickoff. Impressions alone are not a report; they are a receipt.