Here is a pricing anomaly hiding in plain sight. The largest video platform on earth, now watched more on living room televisions than on phones in the US, clears skippable inventory at CPMs a half to a fifth of what "premium CTV" charges, on the same screens, in the same rooms, with measurement that is a generation ahead. The discount exists because of a sorting error in buyers' heads: YouTube got filed under scrappy digital video a decade ago while CTV inherited television's prestige, and media plans still price the filing cabinet instead of the attention.
The skip button is the misunderstood asset. Buyers read it as a defect; it is actually a payment structure. On skippable in-stream you pay when someone watches thirty seconds or clicks, which means the people who skip cost you nothing but a five-second brand flash, a five-second exposure with sound at a price of zero. The economics of that are absurd in your favor: the format charges for demonstrated attention and throws in fleeting reach as a rounding error. Compare that with unskippable CTV, where you pay full freight for every impression delivered to a room that may be empty, a gap I unpacked in CTV sells like TV and measures like display. One format bills you for attention; the other bills you for playback.
The measurement advantage compounds it. YouTube runs on logged-in identity, which makes reach and frequency real numbers rather than device-graph guesses, and brand lift studies are native, cheap, and decently designed. It will never be holdout-grade evidence, the standard I defend in you do not need a data science team to run a holdout, but as platform measurement goes it is the least fictional in video.
Where buyers go wrong is treating YouTube like a performance channel with pictures. Defaulting everything into conversion-optimized campaigns produces the same pathology as everywhere else: the algorithm harvests existing demand, reports gorgeous CPAs, and adds little. The channel's actual edge is efficient, measurable, sound-on reach for brands that cannot or should not pay CTV tolls, plus the one thing nothing else has: creator adjacency at scale, where the content itself confers the trust. The craft requirements are real, hooks built for a skippable first five seconds, creative made for the format rather than repurposed spots, and enough concepts to feed rotation, because the fatigue math from ad fatigue has math, use it applies at video speed.
A starting allocation that has served my clients well: move ten percent of the CTV line into skippable YouTube on the same flight dates, matched geos if you can manage it, and compare cost per reached household and branded-search movement at week six. The comparison rarely needs a statistician. It usually needs someone willing to present it.
The anomaly will not last forever; money eventually finds mispriced attention, and YouTube's living-room share keeps forcing the comparison. Until the filing cabinets update, the trade is available: television's screen, digital's accountability, and a bill that only arrives when someone actually watched.
Quick answers
Is YouTube advertising cheap?
Per second of genuine attention, it is the best deal in video: skippable formats mean you largely pay for people who chose to keep watching, and CPVs commonly land in single-digit cents. Cheap attention still needs creative built for the platform to convert.
What is a good YouTube CPV?
Most advertisers see cost-per-view between roughly $0.02 and $0.15, with targeting and creative pulling more weight than bids. Judge it next to what the view does downstream, branded search and site visits, not against the CPV column alone.