CPM Explained: What You’re Paying For and Whether It’s Worth It

CPM stands for cost per mille, where “mille” is Latin for thousand. It is the price an advertiser pays for one thousand impressions of an ad, meaning one thousand times that ad is loaded and displayed to a user. It is the oldest pricing model in digital advertising and still one of the most widely used, particularly in display, video, programmatic, and social media campaigns.

The formula is straightforward: CPM = (total spend / total impressions) x 1,000. If you spend £5,000 and receive 2,000,000 impressions, your CPM is £2.50. What that number means commercially is a different conversation entirely.

What Does CPM Actually Measure?

An impression is recorded when an ad is loaded. Not when it is seen. Not when it is read. Not when it influences anything. Just when it loads. This distinction matters enormously when you are evaluating CPM as a performance indicator, because the gap between “loaded” and “seen” can be substantial depending on placement, device, and the speed at which users scroll.

Viewability standards, as defined by the Media Rating Council, set a minimum threshold of 50% of an ad’s pixels being visible for at least one second for display, and two seconds for video. Even that is a low bar. An ad that is 50% visible for one second is not an ad that has made any meaningful impression on a human being. But it counts. And it contributes to your CPM calculation.

I spent several years managing large programmatic budgets across retail and travel clients, and the single most common mistake I saw was teams treating CPM as a quality signal. A £1.50 CPM felt like a win. A £12 CPM felt like overspending. In reality, the £1.50 CPM was often buying remnant inventory on low-quality placements with viewability rates below 40%, while the £12 CPM was delivering premium video in-stream with near-total viewability and a relevant audience. The cheaper number was costing more in practice.

If you want a broader grounding in how CPM fits within the wider paid advertising ecosystem, the paid advertising hub covers the full range of channels, pricing models, and strategic frameworks that sit around it.

How CPM Compares to CPC and CPA

CPM is one of three primary pricing models in paid advertising. The others are CPC (cost per click) and CPA (cost per acquisition). Understanding the difference is not just academic. It shapes how you structure campaigns, how you evaluate performance, and what risks you are taking on as an advertiser.

With CPM, you pay regardless of whether anyone clicks or converts. You are buying exposure. With CPC, you only pay when someone clicks, which transfers some of the performance risk to the platform. With CPA, you only pay when a defined action occurs, which transfers even more risk to the platform but typically comes with a higher unit cost and less control.

The right model depends on your objective. If you are running a brand awareness campaign at the top of the funnel, CPM is often the appropriate model because reach and frequency are the actual goal. If you are running a direct response campaign where every pound needs to be accountable to a conversion, CPC or CPA structures give you tighter control. The mistake is using CPM for performance objectives and then wondering why the cost-per-acquisition looks terrible.

I have seen this play out repeatedly with clients who wanted to run “awareness campaigns” but were also quietly hoping for sales. The two objectives are not incompatible, but they require different measurement frameworks. When you try to evaluate a CPM-based awareness campaign using CPA logic, you will almost always conclude it failed, even when it did exactly what it was designed to do.

What Drives CPM Rates Up or Down?

CPM rates vary enormously across channels, audiences, placements, and times of year. Understanding what drives that variation helps you make better buying decisions and avoid paying premium prices for commodity inventory.

Audience targeting is the single biggest driver of CPM. The more precisely you define your target audience, the more you will pay per thousand impressions, because you are competing with other advertisers for the same narrow pool of users. A broad run-of-network campaign might have a CPM of £1 to £3. A tightly targeted campaign reaching senior IT decision-makers in financial services might have a CPM of £25 to £50 or higher on some platforms. You are not overpaying. You are paying for specificity.

Placement quality has a direct impact on CPM. Premium placements, above-the-fold positions, homepage takeovers, and in-stream video inventory command higher CPMs than below-the-fold display or interstitial formats that users ignore or dismiss immediately. The premium is usually justified, but not always. Part of the work is knowing when it is.

Seasonality matters more than most advertisers account for. Q4, particularly October through December, sees CPMs spike across virtually every channel as retail advertisers flood the market. I have seen CPMs on social platforms more than double between August and November on the same audience segments. If you are planning a Q4 campaign without factoring in CPM inflation, your budget projections will be wrong.

Format also drives CPM. Video consistently commands higher CPMs than static display because it demands more attention and delivers more information. Rich media formats sit somewhere in between. The format premium is usually worth paying when your creative is strong enough to use the format well. Paying video CPMs for a static image dropped into a video frame is a waste of money I have seen happen more than once.

CPM Across Different Channels

CPM rates are not comparable across channels without context. A £5 CPM on LinkedIn is not the same as a £5 CPM on a display network, even though the number is identical. The audience quality, intent signals, and commercial context are completely different.

LinkedIn CPMs tend to be among the highest of any platform, often ranging from £20 to £60 or more depending on targeting. The reason is the professional audience data. When you can target by job title, seniority, company size, and industry with reasonable accuracy, you are reaching a commercially valuable audience that B2B advertisers are willing to pay significantly for. The high CPM is the price of that precision.

Meta (Facebook and Instagram) CPMs vary widely but are generally more efficient than LinkedIn for consumer audiences. The platform’s behavioural and interest data is extensive, and the auction is competitive but not prohibitively expensive for most verticals. Video formats on Meta, particularly in-feed and Reels, have seen CPM increases as inventory has become more competitive.

Programmatic display CPMs can look very attractive on paper, sometimes as low as £0.50 to £2, but the quality range is enormous. The programmatic ecosystem includes premium publisher inventory and low-quality long-tail sites in the same auction. Without proper brand safety controls, inclusion lists, and viewability filters, low programmatic CPMs often represent wasted spend rather than efficient reach. Tools like those discussed at Moz’s Google Ads resource highlight how campaign structure and targeting precision affect the quality of what you are actually buying.

Connected TV (CTV) and streaming video CPMs are high, typically £20 to £40 or more, but the completion rates and attention metrics often justify the cost. When someone is watching a streaming service on a television, they are not scrolling past your ad at the speed of a social feed. The format commands attention in a way that most digital formats do not.

The Viewability Problem

Viewability is the most important modifier to any CPM discussion and the one that gets glossed over most often in media planning conversations. If you are buying impressions without knowing the viewability rate of the placements you are buying, you do not actually know what you are paying for.

Industry viewability benchmarks for display advertising have historically hovered around 50 to 60 percent on average, meaning that roughly half of all display impressions are never actually seen. Your effective CPM on a viewable basis could be double what your reported CPM suggests. A £3 CPM with 50% viewability is effectively a £6 CPM for impressions that were actually visible.

When I was overseeing media planning at a performance-focused agency, we started requiring viewability guarantees from publishers before committing to direct buys. It changed the conversations significantly. Some publishers pushed back because their inventory could not meet the threshold. That told us everything we needed to know about where not to spend. The ones who could guarantee viewability above 70% earned a disproportionate share of budget.

Programmatic buying has made viewability measurement more accessible through third-party verification tools, but it has also made it easier to ignore. When you are buying thousands of placements through an automated system, it is easy to look at aggregate CPM and miss the fact that a significant portion of your impressions are loading below the fold on pages no one scrolls to. The data is usually available. The question is whether anyone is looking at it.

Frequency: The CPM Variable Nobody Talks About Enough

CPM tells you how much you paid per thousand impressions. It does not tell you how many unique people those impressions reached, or how many times the same person saw the same ad. Frequency management is the part of CPM-based buying that separates disciplined media planning from media waste.

There is a point at which additional impressions to the same user stop building brand awareness and start generating annoyance. Where exactly that threshold sits depends on the category, the creative, and the platform, but it exists. Serving the same display ad to the same person fifteen times in a week is not fifteen times more effective than serving it once. It is probably less effective than serving it three times and spending the remaining budget on new users.

The challenge is that frequency caps require active management. In programmatic environments, without explicit frequency capping across the campaign, the same cookies or device IDs can be targeted repeatedly across different exchanges. You end up with a headline CPM that looks efficient while the effective reach is far narrower than the impression volume suggests.

Reach and frequency reporting, available on most major platforms, should be reviewed alongside CPM on any brand campaign. If your reach is low and your average frequency is high, you are not buying reach. You are buying repetition to a small audience, which may or may not be what you intended.

When CPM Is the Right Metric to Optimise

CPM is the right primary metric when reach and visibility are the genuine commercial objectives. Not everything in marketing needs to be directly attributable to a conversion. Brand building, category entry point reinforcement, and share of voice maintenance are all legitimate objectives that CPM-based buying serves well. The problem is not the metric. The problem is using it for objectives it was never designed to serve.

New product launches often benefit from CPM-heavy strategies in the early phase. When nobody knows your product exists, you need reach before you can expect response. Trying to run a performance-only campaign for a brand with no awareness is like expecting a search campaign to work when nobody is searching for you yet. The sequence matters. CPM-based awareness creates the conditions in which performance channels can work more efficiently.

I saw this play out clearly when working on a travel client’s seasonal campaign. The team had historically focused almost entirely on paid search, which was efficient but capped. When we added a CPM-based video layer in the weeks before the peak booking window, search performance improved. Not because the video directly drove conversions, but because it built enough familiarity that more people searched by brand name, which converted at a significantly higher rate than generic terms. The CPM spend was not directly attributable to revenue in a last-click model, but the commercial logic was sound.

Understanding how CPM-based buying integrates with search and other performance channels is something the Moz SEO and PPC integration framework addresses well, particularly around how upper-funnel activity influences lower-funnel performance in ways that standard attribution models miss.

When CPM Is Being Used to Hide Poor Performance

CPM can be used as a vanity metric. I have been in agency reviews where the headline number presented to a client was “we delivered 50 million impressions at a £2 CPM.” Impressive-sounding. Commercially meaningless without context. What were the viewability rates? What was the frequency? What audience was actually reached? What happened downstream?

When an agency or platform leads with CPM efficiency as the primary success metric for a campaign that had commercial objectives, that is a red flag. It means the conversation has shifted from “did this work?” to “did we deliver the contracted volume cheaply?” Those are not the same question.

Having judged the Effie Awards, I have seen the other side of this. The campaigns that win on effectiveness are not the ones that led with CPM efficiency. They are the ones that connected media investment to a clearly defined business outcome, whether that was sales, market share, or a measurable shift in brand perception. CPM appears in those cases as an input to the strategy, not the headline result.

The honest version of CPM reporting includes effective CPM on a viewable basis, reach and frequency breakdown, audience composition verification, and some form of downstream signal, even if it is indirect. Brand lift studies, search volume trends, and direct response rate comparisons across exposed and unexposed groups are all imperfect but useful. They move the conversation from “we delivered impressions” to “here is what those impressions may have done.”

There is much more on how to evaluate paid media performance honestly, including the metrics that matter and the ones that obscure more than they reveal, in the paid advertising section of The Marketing Juice.

How to Calculate an Effective CPM

The standard CPM formula is: CPM = (total cost / total impressions) x 1,000. If you spent £10,000 and received 4,000,000 impressions, your CPM is £2.50.

To calculate an effective CPM adjusted for viewability, divide your total cost by the number of viewable impressions rather than total impressions. If only 60% of your 4,000,000 impressions were viewable, your viewable impression count is 2,400,000, and your effective viewable CPM is £4.17, not £2.50. That is a meaningful difference when you are planning budgets and evaluating channel efficiency.

You can also calculate eCPM (effective CPM) when comparing CPM campaigns to CPC or CPA campaigns. Take the total cost of a CPC campaign, divide by total impressions, and multiply by 1,000 to get the equivalent CPM. This allows you to compare the cost of reach across different buying models, which is useful when you are allocating budget across a mixed media plan.

Conversion tracking has become increasingly sophisticated, and platforms have improved their ability to connect impression exposure to downstream actions. Resources like the Search Engine Land overview of Google’s conversion tracking evolution show how far the tooling has come, though connecting CPM-based awareness activity to conversion outcomes remains one of the harder measurement problems in paid media.

CPM in Programmatic Buying

Most programmatic display and video buying is transacted on a CPM basis through real-time bidding auctions. When a user loads a page, an auction runs in milliseconds, advertisers bid for the impression, and the winner’s ad is served. The CPM you pay is determined by the auction dynamics, your bid strategy, and the floor prices set by publishers.

Programmatic CPMs can be set through various bidding strategies. Fixed CPM bids give you cost certainty but may limit delivery if your bid is below the clearing price. Dynamic CPM (dCPM) allows the platform to adjust your bid within a set maximum to improve outcomes. Target CPM strategies optimise toward a cost-per-thousand goal while the algorithm manages individual auction bids.

The programmatic ecosystem is complex, and a meaningful portion of the CPM you pay goes to the supply chain rather than to the publisher whose inventory you are buying. Demand-side platform fees, supply-side platform fees, data costs, and verification costs all come out of the gross CPM. The net CPM reaching the publisher can be substantially lower than what you paid, which has implications for publisher relationships and inventory quality.

Programmatic direct deals and private marketplace (PMP) deals allow you to buy specific publisher inventory programmatically at negotiated CPMs, with better transparency into where your money is going. These tend to command higher CPMs than open auction but deliver better quality control. For brand-sensitive categories, the premium is usually worth paying.

The development of programmatic tools and interfaces has moved quickly. The Search Engine Land coverage of AdWords Editor’s evolution is a useful reminder of how the tooling landscape has shifted, and programmatic buying interfaces have followed a similar trajectory from manual and opaque to increasingly automated and data-rich.

What a Good CPM Looks Like

There is no universal benchmark for a good CPM because the metric is only meaningful in context. A £3 CPM is excellent for broad consumer awareness on a quality publisher network. It is terrible if it represents low-viewability, untargeted remnant inventory that never reaches your audience. A £40 CPM is expensive in absolute terms but may represent outstanding value if it is delivering your exact target audience in a high-attention environment with strong viewability.

The question to ask is not “is this CPM low?” but “what am I getting per thousand impressions, and is that worth the price?” That requires knowing your audience targeting parameters, your expected viewability rate, your frequency cap, and what downstream signal you are using to evaluate whether the impressions are doing anything useful.

When I was running a team managing significant display budgets, we built a simple scorecard for CPM evaluation that weighted viewability, audience match rate, placement quality, and frequency distribution alongside the raw CPM number. It was not sophisticated. It was just a structured way of stopping people from celebrating a low CPM without knowing whether it represented good buying. That discipline saved a meaningful amount of wasted spend over time.

Forrester’s research on audience-based buying and B2B media effectiveness, available through their ABM and audience insights work, reinforces the principle that audience quality and contextual relevance matter far more than CPM efficiency in isolation, particularly for considered purchase categories where the wrong impression is not just neutral but can actively damage brand perception.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions