Yes, but not the way most small businesses picture it. Connected TV advertising no longer means booking a national cable slot through a media buyer — Roku, Amazon, and Google’s own DV360 all sell CTV inventory through self-serve dashboards now, and the real entry point sits closer to €500 than the five-figure minimums agencies quoted three years ago. The catch is that “you can afford it” and “it will beat what you’re already doing on Meta or YouTube” are two different questions, and we see business owners conflate them constantly.
We get asked about CTV almost every time a client’s Meta or Google account plateaus and someone on their team has read that “TV is having a performance marketing moment.” It usually is, for brands with the audience size and creative budget to make it work. For a lot of small businesses, it’s an expensive way to relearn a lesson they already know from YouTube.
What does CTV advertising actually cost in 2026?
CPMs on the major self-serve platforms currently sit in the $20–$60 range, with entry-level, broad-audience placements at the low end and premium inventory — home-screen takeovers, top-rated shows — pushing well past that. Roku doesn’t publish an official minimum spend through Roku Ads Manager, and in practice campaigns can launch around $500. Amazon’s DSP, by contrast, is built for a different scale: self-service accounts are realistically a $10,000+ commitment before the retail-attribution and commerce-signal advantages start paying for themselves. The Trade Desk sits further out still, built around six- and seven-figure annual commitments and a genuinely steep learning curve — not a small business platform.
Google’s DV360 is the middle path most of our clients end up looking at, mostly because it plugs into YouTube CTV inventory and sits next to campaigns a team is already running in Google Ads. If you’ve read our post on YouTube ads production costs, the creative math is similar here: the media budget is rarely the constraint. The 15–30 second video asset is.
The three tiers, in plain terms
- Roku Ads Manager (~$500 to start): the accessible on-ramp. Household and ACR-based targeting, self-serve creative upload, no enforced minimum. Good for testing whether CTV moves the needle at all before committing real budget.
- Google DV360 / YouTube CTV (moderate entry, no hard floor but needs real budget to get signal): the natural next step if you already run Performance Max or YouTube campaigns, since audiences and creative can be shared across formats.
- Amazon DSP ($10k+) and The Trade Desk (enterprise): not small-business starting points. These make sense once you have retail attribution data or a media team dedicated to trading the account daily.
Does CTV actually beat YouTube for a small business?
Not usually, and here’s the honest reasoning: YouTube ads already reach a meaningful share of the “living room screen” audience CTV is sold on, through the same auction and targeting stack you’re likely already using, with lower production stakes because 6-second bumpers and in-feed formats are acceptable there in a way they aren’t on a 55-inch screen. CTV’s real advantage is household-level, shared-screen reach — it’s genuinely good at building broad brand awareness the way a local TV spot used to, and the ACR (automatic content recognition) targeting some platforms offer is more precise than anything traditional broadcast ever had.
Where it tends to disappoint is direct response. CTV inventory is built for viewing, not clicking — there’s no thumb on a remote the way there’s a thumb on a phone — so click-through and last-touch conversion numbers will look worse than Meta or Search almost every time, even when the campaign is doing real work on awareness and assisted conversions further down the funnel.
So when is CTV worth testing?
Our honest answer, after watching a handful of client tests: CTV earns a slot in the budget when three things are true at once. First, you already have (or can afford to produce) at least one broadcast-quality 15–30 second video — repurposed YouTube or Meta video ads with title cards and social-media framing look noticeably cheap on a TV screen. Second, your core paid channels (Meta, Google Search, YouTube in-stream) are already reasonably optimized; if you’re still leaving obvious improvements on the table in your Meta account, that budget will outperform a CTV test every time. Third, you can commit to at least 4–6 weeks and a few thousand euros minimum, because CTV’s real signal — did brand search or direct traffic lift during flight windows — takes longer to read than a Meta A/B test.
| Platform | Realistic entry budget | Best for | Setup effort |
|---|---|---|---|
| Roku Ads Manager | ~€500 | Testing CTV viability, local/niche targeting | Low |
| Google DV360 (YouTube CTV) | €1,500–3,000/mo to get signal | Teams already running Google/YouTube campaigns | Moderate |
| Amazon DSP | €10,000+ | E-commerce brands with retail attribution needs | High |
| The Trade Desk | Six figures annually | Enterprise media buyers with dedicated traders | High |
How we’d structure a first CTV test
If a client asks us to run one, the brief looks almost nothing like a Meta campaign brief. We start with Roku, not DV360 — the lower floor means a bad result costs a few hundred euros instead of a few thousand. We repurpose the best-performing existing video creative rather than commissioning something new for the first test; if repurposed creative performs, that’s the signal to invest in a purpose-built CTV asset. We set the flight for a minimum of four weeks and track branded search volume and direct-traffic lift in the reporting layer we already use for the rest of the paid media budget, rather than judging the channel on last-click conversions it was never going to win on.
What we don’t do is treat CTV as a replacement for underperforming channels. It’s additive reach, not a fix for a Meta account that’s already leaking money. If your core channels aren’t healthy, that’s the €500 better spent first.
The mistakes we see most often
Three patterns come up again and again when a small business runs its own first CTV test without an agency involved.
Using the wrong creative aspect ratio and calling it “close enough.” A vertical Reels asset with burned-in captions and a giant “Swipe Up” arrow reads as an obvious mistake on a television. Viewers notice immediately, and it undercuts the credibility the channel is supposed to buy you. If you don’t have a proper 16:9 asset with no on-screen text overlays competing with a remote-control safe zone, that’s the first thing to fix before spending anything.
Judging the test on last-click conversions after two weeks. CTV is a reach channel measured in lift, not a response channel measured in clicks. Two weeks isn’t enough time for branded search or direct traffic to move in a way you can separate from normal noise, and judging the channel on its worst metric guarantees you’ll conclude it doesn’t work.
Skipping frequency capping. Household-level targeting on a small CTV budget concentrates impressions on a narrow pool of devices fast. Without a frequency cap, the same household can see the same 30-second spot a dozen times in a week, which burns budget on repetition instead of reach. Most self-serve dashboards default this too high — check it before you launch, not after the invoice arrives.
Is CTV worth it for a business under €5,000/month in total ad spend?
Generally, no — not yet. Below that spend level, the budget does more work staying concentrated in one or two channels you can optimize daily than spread thin across a reach channel that needs weeks to show signal. CTV starts to earn a genuine slot once your core paid media spend is in the €5,000–10,000/month range and already performing near its ceiling, because that’s the point where the marginal euro in an existing channel buys less than a first test of a new one.
The honest verdict
CTV in 2026 is genuinely more accessible than the “TV advertising” small businesses remember, and Roku’s self-serve on-ramp in particular is worth a small, time-boxed test if you have decent video creative sitting idle. But it’s a reach-and-awareness channel wearing a performance-marketing costume. Businesses that go in expecting Meta-level click-through and last-touch attribution numbers walk away disappointed; businesses that go in measuring the same lift-based way they’d judge a billboard tend to come away with a genuinely useful addition to the media mix.