Black Friday ad costs: the conversion rate you need for your CPA to hold
When clicks get dearer in the peak weeks, cost per purchase rises unless the website converts more of them. Here is the calculation, a break-even table and where on the site to look.
- CPA is click cost divided by conversion rate. If cost per click rises 20%, purchase conversion rate has to rise 20% in relative terms, for example from 2.5% to 3.0%, for cost per purchase to stay the same.
- The size of the peak-week rise is not settled. Three agency reports on Black Friday 2025 disagree on click costs, so plan from your own account history, channel by channel.
- ROAS follows revenue per visitor. A higher order value can cover part of a dearer click, but only net of the discount and only if the margin still works.
- The lever is on the website. Message match with the ad, first-time visitors from paid social, offer clarity and the mobile path are the first places to check.
CPA is cost per click divided by conversion rate. If your cost per click rises 20% in the Black Friday weeks, your purchase conversion rate has to rise 20% in relative terms, for example from 2.5% to 3.0%, for cost per purchase to stay where it was. The same percentage lift in revenue per visitor keeps ROAS where it was.
How much click costs rise is less settled than most Black Friday planning assumes. The agency reports published after the 2025 peak disagree with each other, and none of them measures the whole market. So this article does not hand you one number to plan around. It gives you the calculation, a break-even table for several click cost increases, and the parts of the website that move post-click conversion when traffic is expensive.
It is written for online stores that run paid ads, so the conversion event throughout is a purchase. Leads and sign-ups follow a different calculation and are left out.
Why CPA rises when click cost rises
Cost per purchase from paid traffic has two inputs: what you pay for a click, and the share of clicks that end in an order.
ROAS = revenue per visitor ÷ cost per click, where revenue per visitor is conversion rate × average order value
Bidding, targeting and creative work on the first input. In the peak weeks that input is harder to hold, because more advertisers bid for the same audiences at the same time. The auction sets what a click costs, and your website sets what a click is worth.
That is why the same campaign, with the same ads and the same landing pages, can report a worse CPA in late November without anyone having made a mistake. Nothing got less efficient. One input moved and the other stayed still. The hub article for this cluster covers the general relationship: why conversion rate lowers the cost of every paid result.
The table below runs the same $30,000 of ad spend through three cases: a normal week, a peak week where the website performs as usual, and a peak week where conversion rate rises enough to offset the click cost. Hypothetical example
| Metric | Normal week | Peak week, same conversion rate | Peak week, CPA held |
|---|---|---|---|
| Ad spend | $30,000 | $30,000 | $30,000 |
| Cost per click | $1.00 | $1.20 | $1.20 |
| Clicks | 30,000 | 25,000 | 25,000 |
| Purchase conversion rate | 2.5% | 2.5% | 3.0% |
| Purchases | 750 | 625 | 750 |
| Cost per purchase (CPA) | $40.00 | $48.00 | $40.00 |
| Average order value | $100 | $100 | $100 |
| Revenue | $75,000 | $62,500 | $75,000 |
| ROAS | 2.50 | 2.08 | 2.50 |
| Revenue per visitor | $2.50 | $2.50 | $3.00 |
With conversion rate unchanged, the same $30,000 buys 5,000 fewer clicks, which means 125 fewer purchases and $12,500 less revenue. CPA goes from $40.00 to $48.00 and ROAS from 2.50 to 2.08. Getting back to $40.00 takes 750 purchases from 25,000 clicks, which is a 3.0% conversion rate.
Note the unit. The lift needed is 20% in relative terms, which here is half a percentage point. Teams that hear “20% more conversions” often picture something much larger than moving from 2.5% to 3.0%.
How much ad costs rise in peak weeks
The honest answer is that it depends on the channel, the vertical and the account, and the published data shows it.
GeistM’s review of Q4 2025 Meta rates, published on 13 January 2026, tracks prices in half-month periods. It reported CPMs up 7% in the second half of October, up 3% in the first half of November and up 11% in the second half of November, the period that contains Black Friday. Over the same three periods it reported CPCs down 14%, up 2% and down 3%. Impressions got dearer while clicks did not, which can only happen when click-through rate rises too. The report does not state how many accounts it covers.
Logical Position’s Cyber Week recap, published on 23 December 2025, reported CPCs up 6% year over year across its client portfolio between Thanksgiving and Cyber Monday, with Cyber Monday up 11% and Google CPCs up 13%. Power Digital’s Cyber 5 analysis, published on 11 December 2025 and based on what it describes as hundreds of e-commerce accounts, reported the opposite on Meta: CPCs down 18% year over year and Google CPCs also down, while Meta CPMs rose 3.1% and TikTok CPCs rose 14.3%.
Three cautions before using any of these figures in a plan:
- They are vendor data. Each agency describes its own clients, with its own mix of verticals and budgets. None is a market sample.
- They measure different things. Two compare the peak with the same days a year earlier. Only GeistM compares the peak with the weeks just before it, which is the comparison that matters for a CPA that was fine in October.
- They disagree on direction. On Meta, one portfolio saw cheaper clicks than the year before while another saw dearer clicks across channels.
So the number to plan with is your own. Pull last year’s cost per click by week and by channel from October to December, and express each peak week as a percentage over the average of the eight weeks before it. That percentage is the input for the next section. If it comes out near zero, most of this article does not apply to you, and that is worth knowing before the budget is set.
The conversion rate that keeps CPA flat
Because CPA is a simple ratio, the break-even figure needs one line of arithmetic:
The table applies it to six click cost increases, using the same baseline as before: $1.00 per click, a 2.5% purchase conversion rate, a $100 average order value, a $40.00 CPA and a ROAS of 2.50. Hypothetical example
| CPC increase | New CPC | CPA if conversion rate stays at 2.5% | ROAS if conversion rate stays at 2.5% | Conversion rate needed | Revenue per visitor needed |
|---|---|---|---|---|---|
| +5% | $1.05 | $42.00 | 2.38 | 2.63% | $2.63 |
| +10% | $1.10 | $44.00 | 2.27 | 2.75% | $2.75 |
| +15% | $1.15 | $46.00 | 2.17 | 2.88% | $2.88 |
| +20% | $1.20 | $48.00 | 2.08 | 3.00% | $3.00 |
| +25% | $1.25 | $50.00 | 2.00 | 3.13% | $3.13 |
| +40% | $1.40 | $56.00 | 1.79 | 3.50% | $3.50 |
Two readings of the table are useful. The middle columns show what happens if nothing changes on the website: at a 25% rise in click cost, CPA reaches $50.00 and ROAS falls to 2.00. The last two columns show the target if CPA and ROAS are to hold: a 3.13% conversion rate, or $3.13 of revenue for every paid visit.
The auction sets what a click costs, and your website sets what a click is worth.
Adjusting for order value
Online stores should read the last column first. ROAS depends on revenue per visitor, not on conversion rate alone, and order value often moves in the peak weeks. If shoppers buy more per order, part of the click cost is already covered:
With click cost up 20% and average order value up 10%, the example store needs 2.5% × 1.20 ÷ 1.10 = 2.73%, not 3.0%. ROAS stays at 2.50. Cost per purchase does not: it rises to $44.00, because fewer purchases are carrying more revenue each. Whether that is acceptable is a question about contribution margin, not about ROAS.
Discounts work in the opposite direction. A sitewide 20% off lowers the value of each order unless baskets grow by more than the discount, so measure order value after the discount, not before. A promotion that lifts conversion rate while cutting order value can hold CPA and still lower ROAS.
What moves post-click conversion in peak weeks
The break-even figure is a target, and the table says nothing about how to reach it. What follows are four places to check between the click and the order during a promotion, in the order a visitor meets them. None of them is a guaranteed lift.
1. Message match between the ad and the page
Peak-week ads make a specific promise: a product, a discount and a deadline. Message match is whether the page repeats that promise in the first screen. A visitor who clicked on “30% off winter boots” and lands on a home page with a general sale banner has to find the offer again, and some will not. The click was paid for at the peak price either way. Our guide to improving Google Ads conversion rate covers the campaign side of the same problem.
- The page does not repeat the ad
- A first-time visitor finds no reason to trust the store
- Terms, end date or exclusions are unclear
- The discount appears late, in the cart
- Slow or long path from ad to paid order
- Performance read as one blended number
2. First-time visitors from paid social
If you raise prospecting budgets for the peak, a larger share of your paid visitors will be new to the store. They arrive from a feed, not from a search, and they know the ad and nothing else. Before they buy they need answers a returning customer already has: when it arrives, how returns work, whether other people were satisfied. A page built for customers who know you is likely to convert this group below your average, and they are the clicks the extra budget bought.
3. Offer clarity
During a promotion the visitor is comparing your offer with several others open in other tabs. State what the discount is, what it applies to, when it ends and what delivery costs, on the product page and again in the cart. An offer that needs a code the visitor has to hunt for, or a discount that only appears at checkout, adds a step at the moment the decision is made.
4. Mobile
Check the device split of your own paid traffic first. Where most paid clicks come from phones, the mobile path from ad to paid order is the path most peak-week clicks take. Walk it on a real phone: page weight with promotional banners loaded, the size of the buy button, the number of fields at checkout and whether wallet payments are on. Report conversion rate and revenue per visitor by device, because a blended figure hides a weak mobile path behind a strong desktop one. The related article on lowering Google Ads CPC and CPA with CRO goes further on post-click fixes.
Common mistake: treating a rising peak-week CPA as a media problem only. The usual response is to cut bids, narrow audiences or pause campaigns, which lowers volume in the weeks when demand is highest. Split CPA into its two inputs first. If cost per click rose and conversion rate held, the campaign is doing what it did in October and the room to act is on the page.
How to run the numbers on your own account
The whole calculation takes an afternoon with last year’s data, and it is best done before budgets are approved:
- Measure your own click cost rise. Take weekly cost per click by channel for October to December last year. Express each peak week as a percentage over the average of the eight weeks before it.
- Record the baseline for paid traffic only. Purchase conversion rate, average order value and revenue per visitor, by channel and by device. Sitewide figures include returning customers and email traffic, and will flatter the number.
- Calculate the target. Baseline conversion rate × (1 + your click cost rise). If you expect order value to change, use the adjusted formula and write down the margin you are assuming.
- Choose the pages. List the landing pages that will take peak-week paid traffic and check each one against the four points above before the first campaign goes live.
- Read it weekly during the peak. Cost per click, conversion rate, CPA, ROAS and revenue per visitor against the target, split by channel and device. If you change the website during the peak, keep a holdout so the effect of the change can be separated from the effect of the promotion.
The fifth step is the hardest. Promotions change demand, so a conversion rate that rises in Black Friday week says little about the website unless there is something to compare it with. This is the part of the equation Pathmonk works on: it reads each visitor’s behaviour during the session, using first-party data and no third-party cookies, and adapts the offer, proof and next step shown on the page, with results measured against a control group. The product page on CRO and personalization explains how it works.
If paid acquisition already feels like it returns less each quarter, the peak weeks are the sharpest version of a year-round pattern, covered in why PPC is killing your marketing.
When this advice does not apply
Holding CPA flat through the peak is one possible objective, and for some stores it is the wrong one. The calculation still helps, because it shows what a different objective costs.
If higher order value, measured after discounts, covers the extra click cost with margin to spare, ROAS holds without any lift in conversion rate. If the goal of the peak is order volume or new customers, a higher CPA can be the plan, provided the ceiling is agreed in advance and repeat purchases support it. And if your own click costs did not rise last year, there is nothing to offset.
Two further limits are practical. The peak is short, so a store with little paid traffic will not see a reliable difference in conversion rate within one or two weeks. In that case fix the known problems on the landing pages before the peak and do not try to test during it. And stock sets its own ceiling: a higher conversion rate on products that sell out by Saturday moves the purchases earlier, not higher.
Key takeaways
- CPA is cost per click divided by purchase conversion rate. When one input rises, only the other can bring the result back.
- The conversion rate needed to hold CPA is the current rate multiplied by one plus the click cost increase: 2.5% becomes 3.0% at a 20% rise.
- Published figures on peak-week ad costs are agency portfolio data and they disagree. Use your own weekly cost per click from last year.
- For an online store, ROAS follows revenue per visitor. Higher order value can cover part of a dearer click, measured after the discount.
- CPA and ROAS can move apart: with higher order value, ROAS can hold while cost per purchase rises.
- Look first at message match, first-time visitors from paid social, offer clarity and the mobile path. None is a guaranteed lift.
- Holding CPA flat is not the right goal for every store. Volume targets, margin from order value and stock limits change the answer.
FAQs on Black Friday ad costs and CPA
What conversion rate do I need if my CPC goes up 20% on Black Friday?
Multiply your current purchase conversion rate by 1.2. CPA is cost per click divided by conversion rate, so a 20% rise in click cost needs a 20% relative rise in conversion rate for cost per purchase to stay the same. A store converting paid traffic at 2.5% would need 3.0%. Use the conversion rate of your paid traffic, not the sitewide figure.
Do ad costs always go up during Black Friday?
Not for every account or channel. Agency reports on the 2025 peak disagree: GeistM reported Meta CPMs up 11% in the second half of November while CPCs fell 3%, Logical Position reported CPCs up 6% year over year across its clients, and Power Digital reported Meta CPCs down 18% year over year across its e-commerce portfolio. All three are vendor data from client portfolios. Check your own weekly CPC from last year before planning.
How do I keep ROAS stable when click costs rise?
ROAS is revenue per visitor divided by cost per click, so revenue per visitor has to rise by the same percentage as the click cost. That can come from a higher purchase conversion rate, a higher average order value, or both. Count order value after the discount, because a promotion that lowers the value of each order raises the conversion rate you need.
Is a higher CPA during Black Friday always a problem?
No. If the goal of the peak is order volume or new customers and you have agreed a CPA ceiling that the margin and repeat purchases support, a higher cost per purchase can be the plan. It is a problem when the store needs CPA or ROAS to hold and nothing on the website changed to offset the dearer click.
What should I fix on my website before Black Friday traffic arrives?
Start with the pages that receive paid traffic. Make the landing page repeat the product, offer and deadline from the ad, answer the questions a first-time visitor has about delivery, returns and reviews, state the promotion terms in plain words, and test the mobile path from ad to paid order on a real phone. Measure paid traffic by channel and device so you can see which of these is losing purchases.
Should I pause ads if CPA rises during Cyber Week?
Check the two inputs first. If cost per click rose and conversion rate held, the auction got more expensive and the question is whether the website can convert more of the clicks. If conversion rate fell, look at the landing page, the offer and the device split before changing bids. Pausing is reasonable when cost per purchase passes the ceiling your margin allows and neither input can be moved in time.
- GeistM: Q4 2025 digital ad rate trends (geistm.com, 13 January 2026)
- Logical Position: Cyber Week 2025 recap (logicalposition.com, 23 December 2025)
- Power Digital: Cyber 5 portfolio analysis 2025 (powerdigitalmarketing.com, 11 December 2025)