CTV advertising is a television commercial delivered over the internet to a TV screen, through a smart TV, a streaming stick or a game console, inside ad-supported services such as Hulu, Peacock, Tubi, Pluto TV and the Roku Channel. It plays full screen and cannot be skipped, like a commercial on cable, but it is bought like a digital ad: you choose the zip codes, the kinds of households and the number of times each home sees it, and you get a report on what was delivered.
That combination is why it has moved from a national-brand tool to something a local business can use. This guide explains how it works, what it costs, where money can leak, how results are measured when nobody clicks a TV ad, and how to decide whether it is right for your business.
The short version
CTV stands for connected TV. The connected part means the television is on the internet, and the ad reaches it through an app rather than a cable box or an antenna. Three things follow from that.
First, the ad can be aimed. A cable spot runs across a whole zone whether the viewers are customers or not. A CTV commercial can be limited to a list of zip codes, a radius around a business, or households that match a description, such as homeowners or people in the market for a vehicle.
Second, the ad is counted. Each play is logged, so the report shows how many times the commercial ran, how many homes it reached, how often each home saw it, and how many plays ran to the end.
Third, the price is set per thousand plays, not per month of airtime, so a campaign can start small, be measured, and grow or stop.
If the terms OTT and CTV are blurring together, the earlier guide on CTV versus OTT separates them. The short answer is that OTT is any video delivered over the internet to any device, and CTV is the part watched on an actual television.
Where TV viewing went
The case for CTV advertising starts with where people watch. Nielsen's monthly measurement of all television use, The Gauge, put streaming at 47.5% of TV viewing in December 2025, the largest share it had ever recorded. Broadcast had 21.4% and cable 20.2%, so streaming was larger than broadcast and cable combined. Netflix alone accounted for 9.0% of all television time that month (Nielsen, 20 January 2026).
Much of that streaming now carries ads. Netflix told advertisers in May 2026 that its ad-supported plan reaches more than 250 million monthly active viewers worldwide, a figure it calculates from members who watched one minute or more of ads in a month, multiplied by estimated household size (Variety, 14 May 2026). Amazon reports an average monthly ad-supported reach of more than 315 million viewers worldwide on Prime Video, from its own internal data (Amazon Ads). Add Hulu, Peacock, Paramount+, Tubi, Pluto TV and the Roku Channel, and the ad-supported streaming audience in the United States is now most of the country.
Advertiser money has followed. The Interactive Advertising Bureau expects digital video, which includes CTV, to pass 80 billion dollars in the United States in 2026, growing 11% in a year and taking more than 60% of all TV and video ad spend for the first time, with CTV itself growing 11% (IAB, 5 May 2026). In the IAB's 2025 survey of 364 advertisers, 36% of those planning to spend more on CTV said the money would come out of linear TV budgets and another 36% said it would come from social media (Marketing Dive, 29 April 2025).
None of those numbers mean a local business should buy CTV. They mean the audience is there, and that the question is whether it can be reached in your area, at a cost that fits, in a way that can be measured.
How a CTV ad is bought and aimed
There are two ways in. A business can buy from a single service through its own self-serve tool, which the guide on advertising on Hulu covers in detail. Or it can buy through a partner that uses a demand-side platform, software that bids on ad breaks across many apps at once. That second route is what most agencies mean by CTV advertising, and it is the one described here.
A campaign is built from four decisions.
Where. The area is drawn first: a metro, a county, a list of zip codes or a radius around a location. The ad is only offered to households inside it. For a business that serves one part of a city, this is the single biggest difference from cable, which sells zones that rarely match a service area.
Who. Within the area, the campaign can be narrowed to households that fit a description, built from the same data sources used in audience targeting for display ads: homeowners, households with children, people shopping for a vehicle or a kitchen remodel, or households that have visited a competitor's location. It can also be left wide, which for many local businesses is the right call, because a plumber's customer is anyone with pipes.
How often. A frequency cap sets how many times one household sees the commercial in a week. Too low and nobody remembers the name. Too high and the budget is spent on the same few homes. Two to four plays a week per household is a common starting point, and the right number depends on the budget and the size of the area.
What. The commercial itself, usually 15 or 30 seconds. It does not need to be a broadcast production. It does need a clear business name, a reason to care, and a way to find you, repeated on screen, because the viewer is on a couch and will search for you later rather than click.
The software then bids on ad breaks inside the apps, one at a time, as they come up for households that match. A campaign across Hulu, Peacock, Tubi, Pluto TV and the Roku Channel is normal. The report shows which apps the plays ran in, which is worth reading, because a plan sold as premium streaming can quietly fill up with low-grade inventory if nobody checks.
What it costs and what the money buys
CTV is priced per thousand completed or delivered plays. Local, premium inventory costs more per thousand than display banners or streaming audio, because a full-screen TV commercial in a household you chose is worth more than a banner nobody looked at.
For a local business, most campaigns start at $2,500 to $5,000 a month in media. The low end works when the area is tight, a handful of zip codes rather than the whole metro. A whole metro, several markets or a statewide plan needs more. Below that range a campaign can run, but it does not reach enough of the right homes often enough to tell you whether it worked, and a test that cannot be judged is money spent for nothing.
Plan on three months. The first month builds delivery and frequency across the area. Judging calls after two weeks tells you nothing, because the households that will remember the name have not yet seen it enough times. A short promotional flight is a poor fit for the channel.
Completion rates are high by the nature of the format. Because the ad cannot be skipped and the television is usually on until the show comes back, somewhere between 90 and 98% of plays run to the end. Peer39, which measures streaming inventory, puts the typical range for non-skippable CTV at 90 to 97% (Peer39 quarterly CTV benchmarks). That number is a check that the inventory is real television, not evidence that anyone paid attention, so treat a completion rate below 90% as a question for the vendor rather than treating one above it as success.
Where CTV can waste money
The weak point of CTV advertising is that the TV cannot tell you whether a person was watching, and the systems that buy the ads cannot always tell whether the device was a television at all. Fraud in this channel means plays logged on devices that are not TVs, or in apps that do not exist, or by software pretending to be a household.
The scale is not small. Pixalate, which audits programmatic traffic, found that 18% of global open-auction CTV traffic in the second quarter of 2025 was invalid, across more than 120 billion impressions reviewed (Pixalate, 27 August 2025). By the first quarter of 2026 the same measure had risen to 25% of global open programmatic CTV traffic, on more than six billion dollars of spend (Pixalate, 25 May 2026).
Verification changes the picture. DoubleVerify reported in May 2026 that fraud rates in CTV campaigns running its protection were below 1%, against nearly 9% in unprotected campaigns, and that it had detected 140% more CTV fraud schemes in the first quarter of 2026 than a year earlier. It also found that buying direct from a publisher is not by itself a defence: in one consumer healthcare campaign bought as a direct deal, 34% of impressions went to bots, and 25% in a large packaged-goods campaign (DoubleVerify, 7 May 2026).
For a local advertiser the lesson is simple. Ask whether the plays are verified by a third party before you are billed for them, which apps the campaign is allowed to run in, and whether you will see the app list in the report. An honest vendor answers all three in a sentence each. The guide to programmatic advertising covers where money leaks across the wider auction, and most of it applies here.
How to tell whether it worked
Nobody clicks a TV commercial, so CTV is measured differently from search or social ads.
The first layer is delivery: plays, households reached, average frequency, completion rate and the app list. That tells you the campaign ran as planned. It does not tell you whether it sold anything.
The second layer is the signal that matters. When a household sees a local business's name on the television several times, the next step is usually a search for that name, a direct visit to the website, or a call. In Google Search Console and Google Analytics, searches for your business name and direct visits are the lines to watch, and they usually start moving in the second month of a campaign. By the third month a lift against the three months before the campaign is fair to talk about, provided the baseline was recorded at the start.
The third layer is attribution: matching the households that saw the ad to visits to a location or to a website, through the same device data used to build the audience. This exists and is useful, and it is also where vendors are most tempted to claim credit for customers who were coming anyway. Ask how the match works and what share of visitors it can see before treating it as proof.
The honest position is that CTV builds the name a household searches for later. The campaigns that look best on paper are the ones where someone wrote down the baseline before the first play.
Is CTV advertising right for your business?
It fits when most of these are true.
- You serve a defined area, whether a metro, a few counties or a region, and you can say which zip codes matter.
- The sale is worth enough that one new customer a month covers a meaningful part of the media cost: home services, remodelling, dealerships, dental and medical practices, financial services, senior living, and higher-ticket retail.
- You can commit to three months, and you have or can make a 15 or 30 second commercial.
- Name recognition is the gap. People who know you buy from you, and the problem is that not enough people in your area know you.
It fits poorly when the offer is a short sale that ends in two weeks, when the area is so small that the budget cannot build frequency, or when the business cannot yet handle more calls. It is also a slower channel than search. Someone typing "emergency plumber" into Google is ready now. CTV is how you become the plumber they already have in mind when the pipe bursts.
For many local businesses the right plan runs CTV alongside search and display rather than instead of them: the television commercial builds the name, search captures the demand it creates, and display retargeting follows the people who visited the site but did not call.
Questions to ask before you sign
- Which apps and services will the commercial run in, and will the report list them?
- Can the area be limited to my zip codes or a radius, with no national audience floor that stops a local campaign from switching on?
- Is every play verified by a third party before it is billed?
- What frequency cap will you set, and why?
- What will the first report show, and when, and what will we compare the third month against?
- Who owns the commercial after the campaign, and what does production cost if you make it?
A vendor who answers those plainly is usually one worth working with. The Connected TV service page sets out how Launch Media answers each of them.
Common questions
Is CTV advertising the same as OTT advertising?
Not quite. OTT is video delivered over the internet to any screen, including phones and laptops. CTV is that same streaming watched on a television. Most local plans buy the TV screen first, because that is where a household sees the name often enough to remember it. The two terms are explained in full in CTV vs OTT.
Can a small business afford CTV advertising?
A single-market test in a tight area starts at $2,500 to $5,000 a month in media, and that is the scale at which the results can be judged. A smaller budget can run, but it is hard to learn anything from it.
Can CTV ads target one part of a city?
Yes. The area can be a list of zip codes or a radius around a location, and that is the main reason local businesses use CTV rather than cable zones.
Do people skip CTV ads?
The ads in ad-supported streaming services are not skippable, which is why 90 to 98% of plays run to the end. Whether the person in the room was paying attention is a separate question, and the fairer measure is whether searches for your name rise over the following weeks.
How long before CTV advertising shows results?
Delivery reports arrive from the first week. Movement in searches for your name and direct visits usually shows in the second month, and a fair comparison against your baseline is possible after three.
Do I need a professionally produced commercial?
No. A clear 15 or 30 second spot with your name, your reason to call and your location, made for the television screen, is enough to start. Many businesses begin with a video they already have.