YouTube Shorts ad cost in 2026

In The Ad Bench's internal working sample for 2026, YouTube Shorts CPM (cost per 1,000 impressions) runs roughly $3–10 with a median around $6. CPC (cost per click) lands $0.20–1.80 with a median near $0.95. CPV (cost per view, counted at 10 seconds or full play, whichever is shorter) sits at $0.003–0.04 with a median around $0.018. CPCV (cost per completed view) runs $0.03–0.08 depending on creative length and hook quality. In the same sample these brackets sit below the TikTok ($4–12 CPM) and Reels ($6–15 CPM) working ranges for comparable audiences; the gap is a sample observation, not a platform rule.

One plausible reason Shorts has run cheaper in that sample is structural. Demand Gen campaigns serve inventory across Shorts, the Discover feed, and Gmail promotions: three surfaces feeding one auction. That extra impression supply can soften auction pressure on any given placement, while TikTok and Reels each concentrate demand on a single primary surface. Shorts CPMs still inflate in Q4 in the working sample, just from a lower baseline.

The four metrics measure different things and should not be optimized in isolation. CPM tells you what the auction costs. CPC measures hook-and-CTA quality together. CPV gauges first-impression retention; CPCV measures whether the full creative holds the viewer. Cheap CPM with expensive CPC can mean you're winning impressions on Discover and Gmail that don't click. See /learn/reels-ad-cost for the same diagnostic framing applied to Meta.

Google Ads' technical minimum budget is low enough to be meaningless for Shorts in practice (confirm the current floor for your campaign type in Google's budget documentation²). Demand Gen learning tends to need substantially more impressions than a single-placement campaign because the system has to allocate across three surfaces and learn which combinations of creative-plus-placement convert for your audience. The Ad Bench's working floor for a Demand Gen learning phase is $20–50/day for at least 7–14 days; this is internal working guidance, not a Google requirement.

Under-budgeted Demand Gen campaigns fail differently from under-budgeted TikTok campaigns. TikTok stalls visibly: delivery drops, frequency caps, the system tells you it's in learning. Google's Demand Gen just delivers inconsistently across surfaces, never accumulates enough signal on any one placement, and you see flat CPA with no improvement after two weeks. The fix is the same as TikTok: concentrate budget on fewer ad groups. In The Ad Bench's experience, one ad group at $50/day usually beats three ad groups at $20/day during learning.

For a deeper breakdown of budget allocation across testing, scaling, and retargeting phases, see /learn/ad-budgeting. Shorts-specific working guidance: budget $1,000–1,800 for a clean creative test across 4–6 variants before drawing conclusions, and don't cross-compare against TikTok results; the audience composition is different enough that creative winners often don't transfer.

Demand Gen campaigns are how you buy Shorts inventory at scale in 2026 (Google migrated earlier video-action formats into Demand Gen; see Google's Demand Gen documentation¹for the current campaign definition). The defining feature is that a single creative pack serves across Shorts, the Discover feed, and Gmail promotional tabs. You upload one vertical video, one square video, several images, and a set of headlines and descriptions, and Google's system mixes them across surfaces and devices. The cost of producing creative amortizes across three placements instead of one, which is one reason Shorts CPMs have run lower in the working sample.

The trade-off is creative control. You cannot easily isolate performance by surface inside a single Demand Gen campaign; the reporting groups everything together by default, and breaking it out requires placement reports and some patience. If you need surface-pure measurement, you can run a Demand Gen campaign restricted to YouTube placements only, but you give up some of the auction efficiency that comes from the wider impression pool.

Full mechanics of Demand Gen campaign structure, asset requirements, and the surface-by-surface optimization loop are in /learn/shorts-demand-gen. The shorthand: build for Shorts first (9:16 vertical, hook in the first 2 seconds), and let the system adapt the secondary assets for Discover and Gmail.

Demand Gen offers four bidding modes in 2026. Maximize Conversions is the workhorse: no target, the system spends the full budget chasing the cheapest conversion. Right choice during learning phase and when you have no historical CPA. Maximize Conversion Value is the same logic optimized against revenue instead of count, which is better for ecommerce with variable AOVs where a $200 order shouldn't be weighted equal to a $20 order.

Target CPA sets the cost-per-action you're willing to pay, and the system optimizes to hit that average. Useful once you have a meaningful volume of recent conversions at a stable cost (The Ad Bench's working rule of thumb is 30+ in the last 30 days). Target ROAS is the same idea for revenue: set a return multiplier (e.g. 350% for 3.5x) and the system bids to hit it. Both targets fail the same way: set them too far from realistic auction prices and delivery collapses because nothing clears the bar.

Practical sequencing, as The Ad Bench working guidance: launch on Maximize Conversions, gather 30+ conversions, then switch to Target CPA at 110–120% of your observed average. Move to Target ROAS only once you have stable revenue tracking and at least 50 purchases in the optimization window. Switching too early typically throws the campaign back into learning.

Campaign objective drives CPM more than most other levers. In The Ad Bench's working sample, Awareness campaigns (pure reach with no click or conversion optimization) run cheapest at $2–5 CPM, because Google serves to the widest available inventory. View and VTR (view-through rate) objectives sit in the middle at $3–7 CPM, with the algorithm steering delivery toward users likely to watch through.

Conversions and Sales objectives are the most expensive at $7–15 CPM, because Google restricts delivery to users with purchase signal in their recent behavior. The counter-intuitive result is the same as on every other platform: the most expensive CPM usually produces the lowest CPA. Brand Lift studies require a minimum spend over the study window (the threshold varies by market and study type; confirm current requirements in Google's documentation) because the methodology needs statistically meaningful exposed-versus-control populations.

CPMs vary widely by vertical because audience size and competitive density vary widely. Approximate 2026 Shorts working ranges from the same internal sample:

• —Beauty / skincare:CPM $4–9. Huge targetable audience absorbs heavy advertiser competition without pushing prices to TikTok levels.
• —Fitness / supplements: CPM $5–10. Ad-policy friction is real on Google but enforcement has felt more predictable than TikTok's in The Ad Bench's experience.
• —SaaS / B2B:CPM $10–20. Cheaper than the TikTok B2B working range, plausibly because Demand Gen taps Gmail promotions where work-context users actually open ads.
• —Fintech:CPM $12–25. Heavy compliance review and incumbent competition keep this the most expensive consumer-facing vertical in the sample.
• —Gaming / apps:CPM $3–8. Shorts has been efficient for mobile app installs in the working sample, particularly for casual gaming.
• —Food / grocery:CPM $3–7. Broad audiences and lower competitive density keep this among the cheapest verticals.
• —Education / courses: CPM $4–9. Strong creative variance; well-scripted ads can beat the working range meaningfully.

For creators monetizing Shorts directly (rather than buying ads against them), the Partner Program revenue picture is very different from long-form YouTube. Shorts revenue is paid from a pooled ad-revenue model: Shorts ad revenue is aggregated, music licensing costs are deducted, and creators are paid a documented revenue share of the remainder by view share. Confirm the current mechanics and revenue-share percentage in YouTube's Partner Program documentation, since program terms can change. In The Ad Bench's working sample, Shorts RPM (revenue per 1,000 views, after YouTube's share) has run roughly $0.15–$0.45, far below typical long-form RPMs; both figures vary widely by niche, geography, and season.

The practical implication for a creator-led monetization plan: Shorts views fund discovery more than income. Illustrative math from the working RPM range: a channel doing 50M Shorts views per month would earn roughly $7,500–22,500 from the Partner Program, meaningful but not life-changing at that scale. Converting even a small share of those viewers to long-form, merchandise, or sponsorships can earn multiples of that from the downstream surfaces. Treat Shorts as a top-of-funnel surface for your own audience, not as the revenue engine.

Full mechanics of Partner Program eligibility, revenue share calculations, and how ad-policy strikes interact with monetization status are in /learn/shorts-monetization-vs-ad-policy.

The Ad Bench methodology: The targets in this section reflect The Ad Bench's current working guidance and internal calibration set. They are designed to frame creative and budget decisions before media spend, not to predict or guarantee campaign performance.

For DTC ecommerce running Demand Gen with Shorts as primary surface, The Ad Bench's working targets are 2–3.5x ROAS on cold prospecting and 4–7x on retargeting. Anything above 3.5x cold is exceptional and often indicates audience saturation is coming, so be ready to refresh creative. Below 2x cold means either creative or unit economics don't support the channel; at 1.5x with 40% gross margin you lose money on every order before fulfillment.

The strategic position Shorts has occupied in the working sample is top-of-funnel reach at lower CPM than TikTok or Reels, with the tradeoff that conversion intent has generally read lower. Viewing behavior and intent vary by account, audience, and market, so treat this as a hypothesis to check against your own data. Operators who do well on Shorts in The Ad Bench's experience either lean into the cheap CPM for awareness and remarket aggressively on other channels, or they craft creative explicitly designed to drive app installs and lead-gen, the two objectives that have monetized the lower-intent audience best in the sample.

For cross-platform planning, compare against /learn/tiktok-ad-cost and /learn/reels-ad-cost. A portfolio pattern that has worked in The Ad Bench's audits: Shorts for cheap reach and lead-gen, TikTok for commerce, and Reels for warm retargeting, not all three for the same objective.

1. Google Ads Help. "About Demand Gen campaigns." Accessed August 18, 2026 (re-verification blocked from this environment on August 24, 2026). support.google.com
2. Google Ads Help. "About campaign budgets." Accessed August 18, 2026. support.google.com
3. The Ad Bench internal working ranges and illustrative math, current calibration set. Reviewed August 24, 2026. No public benchmark source is claimed for cost, CPA, ROAS, or revenue figures.