Every year the same conversation happens in early November: CPMs are up, cost per purchase is up, and somebody asks whether the platform has changed something. It has not. Your competitors have simply all raised their budgets in the same fortnight, and an auction with fixed supply and rising demand does exactly what you would expect. The useful question is not why media got expensive but what you can do about it, and the answer is more than most brands attempt, because almost all of it involves needing fewer impressions rather than paying less for them.

Q4 CPM inflation is the seasonal rise in the cost of advertising impressions caused by advertiser competition during peak retail trading. Inventory is broadly fixed while demand for it rises sharply, so the auction's clearing price goes up. It is a competition effect, not a platform pricing change, and it reverses in January.

What Does It Actually Affect?

Less than the panic suggests, because CPM is only one term in the equation that matters. Cost per acquisition is roughly your CPM divided by the product of your click-through rate and your conversion rate, which means a thirty percent rise in media cost is fully offset by a thirty percent improvement in the combination of creative performance and site conversion. Both of those are things you control.

Peak also brings something that partially pays for itself. The people seeing your ads in late November are further along in their buying intent than the people seeing them in September, so conversion rates typically rise during the same period that CPMs do. That is why plenty of accounts run their most expensive media and their best cost per purchase in the same week, and why judging the quarter on CPM alone leads brands to pull back from spend that was working.

The number that decides it is contribution margin per order, which is the whole argument of our post on the Q4 profit trap.

How Should You Forecast It?

From your own account, not from an industry figure. Pull CPM by week for the past two years across your main channels, overlay your cost per purchase and conversion rate, and you have a forecast that reflects the specific auctions you compete in rather than an average across categories that behave nothing like yours. Benchmarks are useful for arguing with a board and close to useless for planning a budget.

Do the same for the shoulders. Many brands find their October and early December CPMs are materially below the peak weeks while intent is already elevated, which is the single most actionable pattern in the data and the basis for most of what follows.

What Are the Levers?

Eight, roughly in order of how much they return.

Buy the audience early. October impressions cost less and build the retargeting pools, subscriber lists and creative learnings you will monetise in November. This is the closest thing to arbitrage available in Q4, and it requires nothing but deciding in September to fund it, which is why it heads the allocation advice in our budget pacing post.

Improve creative. Better creative raises click-through and conversion, which reduces the number of impressions each order requires, which is mathematically identical to a CPM reduction. It is also the only lever that keeps working as competition rises, the case in our Q4 creative strategy post.

Fix conversion rate before peak. A site improvement made in September applies to every expensive impression you buy in November, so pre-peak checkout and page work has a multiplier that the same work in January does not, which is why it appears in our checkout optimization guide.

Shift revenue to owned channels. Email and SMS cost essentially nothing per send, so every point of Q4 revenue that comes from your own list is a point you did not buy in a bidding war. Brands with well-built flows and a list grown through October are structurally advantaged in November, which is the mechanic behind our email marketing guide and SMS guide.

Protect returning-customer revenue. A repeat purchase requires no acquisition spend at all, so the retention work that looked optional in Q2 becomes a direct hedge against auction prices in Q4, the argument in our post on acquisition without retention.

Raise average order value. If each order is worth more, you can afford a higher cost per acquisition at the same margin, which effectively buys you room in the auction. Bundles and thresholds do this without deepening the discount, per our AOV guide.

Send better value signals. Automated bidding spends against the values you report, so accounts that pass accurate, differentiated conversion values compete more intelligently in an expensive auction than accounts sending a flat purchase event, covered in our first party data guide.

Test lower-competition channels. Retail advertiser concentration is uneven, and channels with less of it have historically offered relief during peak, Pinterest being the clearest recent example in our own client work, per our Pinterest ads guide. Test these in October rather than discovering them in December, since a new channel needs learning time you will not have at peak.

When Should You Actually Stop Buying?

When contribution margin per order sits below your floor for more than a couple of days at the current spend level. Not when CPM crosses a number that feels high, and not when platform ROAS dips, because both can move against you while the quarter remains profitable. The reverse also holds: an uncomfortable CPM with healthy margin per order is a signal to keep going, and brands that pull back on CPM alone routinely leave profitable volume on the table in the one period when volume is available.

If you do need to pull back, cut spend and keep the offer rather than deepening the discount to rescue efficiency. Discounting into an expensive auction attacks the wrong term in the equation and worsens the cohort you acquire, which then shows up in January, as we cover in post-holiday retention.

The Bottom Line

You cannot change the price of an impression, so change how many you need. Buy audience in October while it is cheap, make creative and conversion rate do more work per impression, move revenue onto channels you own, protect repeat purchase, and judge the whole thing on contribution margin per order rather than on a CPM figure that was always going to rise. Handled that way, an expensive auction is a planning constraint rather than a crisis.

Our Biggest Q4 Guide Lands Soon

The full Q4 profit playbook is nearly finished, with the CPM forecasting model and the efficiency levers we work through on client accounts before peak. Our newsletter list gets it first. Join at webtopia.co/newsletter to be first to get it, plus the platform changes that matter each Tuesday in Beyond the Clicks.

Want to Cut Your Cost Per Order Before Peak?

Book a call and we will work through creative, conversion rate and signal quality while media is still cheap, part of how we work as an ecommerce paid media agency and an ecommerce marketing agency.

Frequently Asked Questions

Why do CPMs rise in Q4?

Advertising inventory is broadly fixed while demand rises, since nearly every retail advertiser raises budgets in the same six weeks. Auctions allocate limited impressions to the highest bidders, so the clearing price goes up.

How much do CPMs typically rise during peak trading?

It varies by category, market and audience, so use your own weekly CPM history from the last two years rather than an industry average.

Does higher CPM mean higher acquisition cost?

Not necessarily. Cost per acquisition is CPM divided by click-through rate times conversion rate, so better creative and a better converting site can offset rising media cost, and peak intent lifts conversion rates too.

What can you do about rising Q4 CPMs?

Buy audience earlier, improve creative and conversion rate, shift revenue to owned channels, protect returning-customer revenue, raise AOV, send better value signals and test lower-competition channels.

Should you pause advertising when CPMs get too high?

Only when contribution margin per order falls below your floor and stays there. Peak CPMs come with peak intent, so expensive media can still be profitable.

Which channels are cheaper during Q4?

Generally those with less retail advertiser competition, which has recently included Pinterest and for some categories Reddit and YouTube. Owned channels are always cheapest, which is why building the list in October is the strongest hedge.