What YouTube Ads Actually Cost: CPV and CPM Benchmarks Worth Quoting

A client asks what YouTube ads cost and most agencies reach for a number they half remember from a blog post two years old. That gap between a guessed number and a sourced one is where quotes either scare a client off or leave money on the table once the campaign actually runs. YouTube pricing is an auction, so there is no fixed rate card, but the published benchmark ranges are specific enough to build a real quote instead of a placeholder.

What CPV and CPM are actually paying for

Cost-per-view (CPV) is Google's own definition for TrueView bidding: an advertiser sets the amount they are willing to pay for a view, and a view counts differently by format. For a skippable in-stream ad, a view registers at 30 seconds watched (or the full video if it's shorter) or on an interaction, whichever comes first. Non-skippable and bumper formats do not use CPV at all, they run on target CPM, meaning the advertiser pays per thousand impressions regardless of whether the viewer watched two seconds or the whole thing. Knowing which pricing model applies to which format is the first thing that has to be right before a quote means anything.

The ranges to actually quote from

WordStream's guide to advertising on YouTube puts typical CPM at $4 to $10 for most advertisers, with a realistic starting Max CPV bid of $0.05 to $0.10 for a new campaign and roughly $2,000 to reach 100,000 views at that rate. Store Growers' 2026 YouTube benchmark data breaks that average out by format, and the spread between formats is wide enough that quoting a single blended number to a client is close to meaningless:

• Skippable in-stream: roughly $5 to $10 CPM, $0.05 to $0.10 CPV
• Non-skippable in-stream: roughly $6 to $10 CPM (impressions only, no CPV)
• Bumper ads (6 seconds): roughly $3.24 to $4.37 CPM
• In-feed / Discovery ads: roughly $3 to $8 CPM, charged only on a click to watch
• YouTube Shorts: roughly $4 CPM, $0.10 to $0.30 CPV
• Connected TV placements: roughly $8.72 to $10.01 CPM

Why vertical and device swing the number further

Averages hide the real spread. Digital Applied's 2026 industry breakdown shows TrueView CPM running as high as roughly $19 in legal services and finance, against roughly $7 to $8 in retail and CPG, a gap driven by how much competition exists for that audience in the auction rather than anything about YouTube's pricing itself. View rate follows the same pattern: Store Growers puts the average TrueView completion rate at 31.9%, but desktop consistently outperforms mobile and tablet, and the gap between the strongest and weakest month of the year in their data is wide enough that a quote built on a single annual average will be wrong in both directions depending on when the flight actually runs.

Seasonality moves the number as much as vertical does

Cost also swings hard by month, and a quote that ignores the calendar will miss on both sides of the year. Store Growers' seasonal breakdown shows CPM dropping to roughly $1.98 to $2.50 in January and February, climbing to $6.00 to $6.33 in April and May, dipping again to $1.76 to $3.00 over the summer, and rising through the fourth quarter to a $5.70 average with peaks near $6.93 in December as retail and other seasonal advertisers compete for the same inventory. A flat annual quote for a client planning a Q4 launch is going to run over budget against these numbers, and that same flat quote applied to a Q1 flight will look artificially expensive next to what actually gets spent.

Which bid strategy actually sets the number a client sees

The benchmark ranges above assume manual Max CPV bidding, which is only one of the available strategies. Google's format guide also lists Target CPM, Target CPA, and Maximize conversions as alternatives for skippable in-stream campaigns, each of which hands bid-setting to Google's automated systems rather than a manually set ceiling. A campaign optimizing for conversions under Target CPA will often show a different effective CPV than a manually bid campaign chasing raw view volume, because the two are being priced against different outcomes entirely. Quoting a client on CPV alone only makes sense when the actual campaign goal is views; once the goal shifts to leads or sales, the relevant number to quote is cost per conversion, not cost per view, even though the underlying auction is the same one.

Turning ranges into a real quote

The workable approach is to quote a range tied to format and vertical, not a single number pulled from a homepage stat. Start a new skippable in-stream campaign at a Max CPV bid in the $0.05 to $0.10 band, budget in the $10 to $50 daily range while the algorithm has room to optimize, and set client expectations using the $2,000-per-100,000-views baseline as a floor rather than a promise, since a competitive vertical will push that up. For a client evaluating CPC-based channels against video, that comparison needs its own framing too, since YouTube is priced and measured differently than search, and a ROAS target built for search intent does not transfer cleanly to a discovery-oriented video buy without adjusting the model first. A three-tier proposal, one figure for awareness-priced bumper and Shorts reach, one for consideration-priced skippable in-stream, one for conversion-priced Target CPA, reads as informed rather than padded, because each tier is tied to a distinct pricing mechanism instead of an arbitrary markup on a single blended number.

Where margin fits without inflating the quote

The benchmark ranges above are media cost, not the number that goes on a client invoice, and the two should stay visibly separate rather than blended into one padded figure. A client who later sees a WordStream or Google benchmark and compares it to what they were quoted should find the difference explained by a clearly stated management fee, the same markup versus margin distinction that governs any resold media spend, not by an unexplained gap between the benchmark and the invoice. Quoting media cost and management fee as two separate line items, rather than one blended number built to obscure the split, is what keeps the pricing conversation defensible if a client ever pulls the same benchmark data this article is built from.

A worked example: turning a $12,000 monthly budget into a format mix

The ranges above are easier to apply once they are run through an actual number. Take a $12,000 monthly budget for a client who wants both reach and a measurable conversion signal. A workable split puts roughly $2,000 into bumper ads at the $3.24 to $4.37 CPM range from Store Growers' data, which buys somewhere in the neighborhood of 460,000 to 620,000 impressions purely for frequency and recall. The bulk of the budget, say $8,000, runs on skippable in-stream under a drive-conversions objective at the $0.05 to $0.10 Max CPV band, landing between 80,000 and 160,000 views depending on how competitive the auction is for that vertical. The remaining $2,000 goes to Shorts at the $0.10 to $0.30 CPV range to pick up the mobile-first segment the in-stream buy under-reaches, producing somewhere between 6,600 and 20,000 views on its own. None of those figures are a promise, they are the same benchmark ranges from above run through a specific number, which is what turns a vague answer about what YouTube costs into something a client can actually budget against.

Reading the first two weeks of live data against the quote

The quote is built from benchmark ranges, but the number that actually matters is what the account does once it is live, and the first two weeks of spend is where that gets checked. If the campaign's actual CPV is running meaningfully above the quoted band once the algorithm has had a normal learning period, the fix is rarely to change the bid strategy mid-flight; it is to check whether the targeting is too narrow for the budget, whether the creative's first five seconds are giving viewers a reason to skip immediately, or whether the vertical genuinely sits at the expensive end of the range this article already flagged. View rate is the other number worth watching daily rather than monthly: a skippable in-stream campaign tracking well under the 31.9% completion average Store Growers reports is a sign the creative or targeting needs attention before the flight has spent through a meaningful share of budget, not a data point to revisit at the end of the month when the spend is already gone.

What Conduit's own account data shows against the published ranges

Published benchmarks describe the market; an agency's own account history describes what a specific piece of creative and a specific audience actually did, and the two numbers are not always close. Conduit's own Manta case study is worth citing directly to a client asking why a quote might beat the WordStream or Store Growers averages: a media mix that included YouTube alongside Meta, TikTok, Snapchat, and X ran to a blended cost per view of $0.10 across more than 772,000 views, sitting at the efficient end of even the tightest CPV band quoted above, achieved through creative built and tested against each audience segment rather than one video run everywhere. That is not a claim that every account lands at $0.10, vertical, creative quality, and auction competition all still apply, but it is the kind of account-specific proof point a published industry average can never provide on its own.

None of these ranges substitute for pulling an agency's own account history once campaigns are live, but they are close enough to build a defensible first quote instead of a guess. Conduit's white label YouTube advertising team runs these campaigns daily across enough accounts to know which of these ranges actually holds for a specific vertical before the first dollar is spent, and prices client proposals off that account history rather than a published average.