TV Ads vs YouTube Ads: How to Choose
TV buys mass reach. YouTube buys targeting you can measure. And most YouTube watching now happens on the TV. How a producer picks between them.
If you are weighing a television campaign against YouTube, here is the decision in one line: television still buys the widest reach money can buy, and YouTube buys attention you can point at a specific person and then count. I have produced spots that ran in both places for the same brand in the same quarter, and the brief for each read like a different craft. This guide walks through where the two split, where they now overlap, and how I help clients pick a screen without burning a media dollar.
TV buys reach. YouTube buys aimed, measurable attention.
A television buy puts one message in front of a large, loosely defined audience at a scheduled moment. You are renting a slice of a program’s viewers and trusting that enough of them look like your customer. YouTube starts from the opposite end. You describe the person you want to reach, an auction serves your ad to people who fit, and you watch what each view leads to next.
Neither is better in the abstract. They answer different questions. Here is the contrast I sketch before any planning:
| Television | YouTube | |
|---|---|---|
| Bought by | Program and an audience estimate | Auction, per view or per impression |
| Targeting | Demographic and daypart | Search history, interests, life events, custom intent |
| Creative slot | 15, 30, 60 seconds, fixed | Six-second bumper up to longer skippable |
| Measurement | Modeled reach (GRPs) | View rate, click rate, conversions, close to live |
| Strongest at | Broad awareness and authority | Response you can trace to a sale |
If you want the wider map of where these formats sit, our rundown of the types of video advertising lays out every placement alongside these two.
Most YouTube watching now happens on the television
This is the shift that makes the old TV-versus-YouTube framing feel dated. In the US, around 60% of YouTube watch time now happens on a TV set, up from roughly 40% in 2022, according to Think with Google. Nielsen’s The Gauge put YouTube at 12.5% of all television viewing in May 2025, the largest share of any single streamer and ahead of Netflix. So when a buyer pictures “a YouTube ad,” they may really be picturing someone on a couch, ten feet from a 55-inch screen, watching through the living room television.
That convergence changes two things I care about as a producer. A YouTube ad on a connected TV plays full screen with the sound on, so it behaves far more like a broadcast spot than a phone ad squeezed between Shorts. And the money is following the eyeballs: connected TV ad commitments are now on track to pass primetime linear commitments for the first time, while linear ad spend is set to fall more than 11% in 2026, per eMarketer. The screen you grew up calling “the TV” is increasingly a YouTube screen.
Streaming passed broadcast and cable combined for the first time. YouTube alone was 12.5% of all TV viewing, the largest share of any single streamer.
Source: Nielsen, The Gauge
How each platform finds your audience
Television targeting is program-centric. You buy time in a show your customer is likely to watch, then accept the spillover. If you sell high-end grills and buy a spot in a cooking competition, you also pay to reach apartment renters with no yard and the people who ordered takeout that night. It works. The waste is the toll you pay for the reach.
YouTube targets the person rather than the program. Through Google’s data you can reach people by recent searches, by in-market signals (Google can see who is actively comparing products in your category right now), by life events like a recent move, or through a custom audience you build from keywords and competitor URLs. The grill brand can reach someone who searched “offset smoker vs pellet grill” last week, in the moment they are watching a barbecue video. That precision changes what the ad has to accomplish, which is where creative comes in.
The creative each format demands
Television creative is built for a fixed, uninterrupted slot. You get 15, 30, or 60 seconds that nobody can skip, so the craft is classic: set up, build, pay off, land the brand. Production values matter because your spot sits next to other polished spots and a weak one shows. Producing a broadcast-grade commercial is its own discipline, and if you are costing that out, our notes on TV commercial production cover the moving parts.
YouTube hands you a menu instead of a single slot. The main three, defined the way Google Ads documents them:
- Skippable in-stream. The viewer can skip after five seconds, and roughly three in four do, so those first five seconds are the whole game. You pay when someone watches 30 seconds (or the full ad if it is shorter) or clicks.
- Bumper. Six seconds, no skip, sold on impressions. One idea, one image, then it is gone. These are close cousins of the short video ads that run across social feeds.
- Non-skippable in-stream. Usually 15 to 30 seconds of guaranteed view, which you earn the right to run by having something worth 30 uninterrupted seconds.
The mistake I see most often: a brand takes its finished 30-second TV spot, uploads it as a skippable YouTube ad, and leaves the hook at second 20 where a broadcast audience expects it. On YouTube that audience is long gone. If you are unsure how long each version should run, we broke down how long YouTube ads can be by goal.
What you can actually measure
Measurement is where the two split the hardest. For decades, television has been counted with Gross Rating Points, a modeled estimate of how much of your target audience probably saw the ad. It is directional and fine for planning reach, but it is a projection, not a headcount. If the vocabulary of reach and frequency is fuzzy for you, we explain reach and impressions in plain terms.
YouTube reports actual behavior. You see the view rate (how many people watched a meaningful chunk) and the click rate, plus conversions tied to what viewers did on your site afterward, close to live. You can also read audience retention to find the exact second people dropped off, which is the sharpest creative feedback loop I know of. It tells you whether the hook, the middle, or the offer is losing them. When a client asks whether the work is paying back, YouTube hands you a number, and turning that number into a revenue story is the job of our guide to video marketing ROI.
One thing worth keeping straight: television still moves the market in ways its own dashboard cannot see. Brands routinely watch branded search and direct traffic spike the day a spot airs, the so-called halo effect. It is real, and it resists clean attribution.
What it costs to get in the game
Television asks for money before anything runs. You commit to a media buy, often months ahead of air, and you produce to a standard that survives a big screen. That combination is why national TV has long belonged to large advertisers, though local and off-peak buys pull the entry point down for smaller companies.
YouTube runs on an auction. You can start a test with a modest amount, pay per view or per thousand impressions, read what the numbers say, and pour more into the winners. A small company can put real video advertising in front of the right people without a seven-figure upfront commitment. That accessibility is the single biggest reason so many brands who never touched broadcast now run video at all.
How I decide: TV, YouTube, or both
When I sit with a brand, the choice usually comes down to what they need to happen next.
Lean television-first when the goal is broad recognition, fast. A launch that has to feel big, a category where being seen as the established name is half the battle, or an audience that skews older and lighter on digital. One spot in the right program can still make a brand feel major overnight.
Lean YouTube-first when you need response you can trace. Sign-ups, sales, qualified leads, or a narrow audience you can define and follow. Direct-to-consumer brands live here because every dollar ties back to an outcome.
More and more, the answer is both, in sequence. Use the reach surface (increasingly YouTube on connected TV, sitting beside or in place of linear) to make a lot of people aware, then use YouTube’s targeting to re-reach the ones who leaned in and walk them toward buying. The two stopped being rivals the day most YouTube viewing moved onto the television.
Where this leaves you
The platforms are converging, but the job under them has not moved. Someone has to decide what the first five seconds say, whether the story earns a full 30, and how one idea flexes from a six-second bumper into a broadcast spot. That judgment is the actual work. If you would rather have a standing team hold that judgment across every cut and placement, that is the kind of embedded creative team Moonb is built to be. Whichever way you go, start from the outcome you need and let it pick the screen.
Frequently asked questions
You can upload it, but I would not run it as-is. A broadcast spot saves its reveal for the end, and on a skippable YouTube placement most viewers are gone by second five. Recut a version that front-loads the hook and the brand, keep a six-second bumper cut for reach, and make a square or vertical variant for phone placements. Same idea, three deliverables shaped to how each surface actually plays.
Sequence them. Use the broad-reach surface (a linear buy, or YouTube on connected TV) to introduce the brand to a large audience, then retarget the people who watched a meaningful chunk with a second YouTube ad that pushes a specific action. The first ad earns recognition, the second converts the ones who leaned in. Running them in isolation is the common miss; the lift shows up when the awareness pass feeds the response pass.
It is a YouTube ad delivered on connected TV, and you buy it through Google Ads, not through a network or station. It matters because you keep YouTube's targeting and near-live measurement while gaining the full-screen, sound-on attention of the living room. When you plan reach, count YouTube-on-TV as its own line rather than folding it into a linear estimate, since the two are bought and measured in completely different systems.
