Amazon Cost Per Click (CPC) in 2026 Average Rates and How to Lower Them

Three CPC articles, three different platform averages: $1.22, $1.18, $0.99. None of them tells you whether your own number is a problem, because none of them knows your category, your ad mix, or your conversion rate.

This page does two things the benchmark tables do not. It explains why those numbers disagree and how to read them. And it makes the case that CPC is the wrong thing to optimize, because you can always lower it by bidding less and lose money doing it.

There is no good CPC in the abstract. A good CPC is one where your ad cost per sale stays below your break-even.

Ad cost per sale = CPC divided by conversion rate

A $1.00 click at a 10% conversion rate means ten clicks per sale, so $10.00 of ad spend per unit. The same $1.00 click at 20% costs $5.00 per sale. The click price did not change. The economics did.

Whether that number works depends on your margin, which is where break-even ACoS comes in. Our guide to what a good ACoS looks like covers that math, including how to set a target from your own contribution margin rather than a category average.

Published platform averages sit roughly between $0.70 and $1.30. Ad Badger reports $1.22 as of September 2026, up about $0.10 year over year, with a low of $1.02 in October 2025 and a high of $1.32 in June 2026. Epinium puts the 2026 average near $1.18. SalesDuo gives a range of $0.70 to $1.50.

By ad format (Keywords.am, 2026):

Format	Reported CPC Range
Sponsored Products	$0.85 to $1.30
Sponsored Brands	$1.10 to $2.50
Sponsored Display	$0.80 to $1.60

Sponsored Display is where sources diverge hardest. Epinium reports roughly $3.72 after a 49% year-over-year rise. SalesDuo reports $0.60 to $1.30 for the same year. That is not a rounding difference, and neither figure should be used to judge your account.

By category, with the numbers that matter.

CPC alone is a price list. Divided by conversion rate, it becomes a decision tool. Category CPC and conversion rates below are from Ad Badger's September 2026 data. Clicks per sale and ad cost per sale are calculated from those two figures (clicks per sale = 1 divided by conversion rate; ad cost per sale = CPC divided by conversion rate).

Category	Avg CPC	Avg CVR	Clicks per Sale	Ad Cost per Sale
Books	$0.38	18.0%	5.6	$2.11
Clothing & Apparel	$0.72	8.6%	11.6	$8.37
Sports & Outdoors	$0.82	10.3%	9.7	$7.96
Electronics	$1.45	9.5%	10.5	$15.26

A Books click costs a quarter of an Electronics click. The sale costs a seventh as much. Cheap clicks in a low-converting category are not cheap, and an expensive click in a high-converting category can be the best money in the account.

The trend.

Keywords.am reports CPCs climbing from $0.89 in 2023 to $1.21 in early 2026, about 35% in three years. Direction matters more than the decimal: click prices are rising market-wide, and no bidding tactic reverses that.

Why These Numbers Disagree

Every provider above is measuring something slightly different:

• Different panels. Each reports its own customer base, skewed by category and advertiser size.
• Different format mixes. Blending Sponsored Brands and Display into one average raises it. Sponsored Products alone is lower.
• Different date ranges. CPC is seasonal. A figure covering Q4 is not comparable to one covering spring.
• Different marketplaces and currencies. A global average and a US average are different numbers.

Use published averages to sanity-check direction, not to judge your account. Your own trailing 90-day CPC, split by ad type and category, is the only benchmark that should drive a decision.

You set a maximum bid rather than a fixed CPC. Amazon runs an auction for eligible ad opportunities, and factors including your bid and ad relevance affect whether your ad is shown. Dynamic bidding and placement adjustments can also change the bid Amazon enters into a particular auction. Your realized CPC therefore does not simply equal the maximum bid you entered.

A simplified illustration.

Three advertisers compete for one placement with broadly comparable relevance. A bids $2.00, B bids $1.40, C bids $1.10. A wins and pays something closer to B's bid than to its own maximum. This is a teaching example, not a model of the full auction: relevance, predicted conversion likelihood, dynamic bidding, and placement adjustments all influence both who wins and what they pay.

Placement and audience adjustments compound.

Amazon's documentation on adjusting bids by placement and audience shows the adjustments applied in sequence, so a $1.00 bid with a 50% audience adjustment, a 100% Amazon Business adjustment, and a 50% top-of-search adjustment produces a final bid of $4.50. Stack that with an up-and-down bid strategy and the theoretical ceiling climbs again. Calculate your maximum possible CPC before enabling multiple adjustments together.

Bidding strategies (confirm current names in the ad console, since Amazon updates them):

• Dynamic bids, down only. Amazon lowers your bid when a conversion looks unlikely.
• Dynamic bids, up and down. Amazon can raise your bid for likely conversions and lower it otherwise, which increases volatility in realized CPC.
• Fixed bids. Your bid is used as stated, adjusted only by placement and audience settings.
• Rule-based bidding. You set performance guardrails and Amazon adjusts bids to stay within them. This is the option most sellers have never opened.

One more distinction: Sponsored ads are priced per click, while DSP and some Sponsored Display placements are priced per thousand impressions. They are not comparable metrics. Our guide to DSP minimum spend covers impression-based buying.

Before trying to lower CPC, identify why it is high. The causes below are diagnostic signals, not fixes; Section 05 covers what to do about each one.

1. Category competition. Largely outside your control. In the benchmark data above, Electronics CPC is several times higher than Books, illustrating how dramatically the market price of a click can vary by category.
2. Head terms instead of long-tail. Broad category terms cost more than specific long-tail equivalents in the same category, and typically convert worse.
3. Weak conversion rate. Relevance and predicted conversion likelihood can affect auction behavior, particularly when dynamic bidding is enabled. Even where CPC itself does not move, weak conversion makes each paid click less economically productive. A clicks-without-sales problem is usually a listing problem.
4. Targeting breadth. Broad match buying discovery traffic at prices you would pay for intent.
5. Bid strategy mismatch. Up-and-down bidding in a volatile category raises realized CPC, especially stacked with placement adjustments.
6. Seasonality. Published monthly data shows meaningful within-year swings, from $1.02 in October 2025 to $1.32 in June 2026 on Ad Badger's figures. Compare like periods.
7. Format mix. Adding Sponsored Brands or Display raises a blended account average without anything going wrong.

Your CPC may not be the problem.

If ad cost per sale looks worse without CPC moving, check your measurement before your bids. Attribution windows and blended reporting can make a stable account look like it is deteriorating, as our attribution guide explains.

Ranked by effect on cost per sale, not by how easy they are:

Lever	Effect on Cost per Sale	Effort
Fix the listing (images, title, price, reviews)	Highest. Conversion rate divides into every click	High
Negative keywords and search term pruning	High. Removes clicks that never convert	Medium, ongoing
Move proven terms to exact match	High. Buys intent instead of discovery	Medium
Placement bid adjustments	Medium. Top of search costs more and converts more	Low
Bid strategy selection	Medium. Changes volatility more than average price	Low
Dayparting	Low to medium, category dependent	Medium
Cutting bids across the board	Lowers CPC, usually raises cost per sale	Low

Fix the Listing First

Conversion rate is the denominator under every click you buy. Take the Electronics example: at 9.5% conversion and a $1.45 click, ad cost per sale is $15.26. Move conversion to 12% and it falls to $12.08, with no bid change at all. Images, title, price, and review quality do more for ad economics than any bidding adjustment.

Start with the ASINs carrying the most ad spend, not the ones with the worst conversion rate, because that is where a percentage point is worth the most money. Our listing optimization checklist covers the work itself.

Prune Search Terms on a Schedule

Pull the search term report weekly. Add as negatives any term with meaningful clicks and no sales, using a threshold set from your own conversion rate rather than a generic number: if you convert at 10%, a term with 40 clicks and no orders is a clearer signal than one with 8.

Amazon's targeting guidance suggests evaluating performance after at least 20 clicks before adding negative targeting, and letting negative keywords run for two weeks or more before making further strategy changes. Resist the temptation to prune daily, which mistakes noise for a pattern.

Graduate Proven Terms to Exact Match

When a term converts consistently in broad or phrase match, give it its own exact-match ad group with its own bid. You stop paying discovery prices for traffic you have already qualified, and you can bid the term to its actual value instead of an average across the ad group. Add the graduated term as a negative in the broad campaign that found it, or the two will compete against each other in the same auctions.

Adjust by Placement, Not Just by Keyword

Pull the placement report. If top of search converts well above your other placements, a placement multiplier is more precise than raising every bid, because it buys the premium position only where the data supports it. If top of search converts worse than rest of search for a given campaign, the multiplier comes down.

Start from zero adjustment until you have placement-level data, then move in increments. Remember the compounding math above: adjustments stack with each other and with dynamic bidding.

Choose the Bid Strategy Deliberately

Most accounts run whatever strategy was selected at campaign creation and never revisit it.

• Down only suits campaigns where you want a predictable ceiling, including new campaigns still gathering data and any campaign where cost control matters more than volume.
• Up and down can work on proven, high-converting terms where winning the better placement is worth a higher price, and is risky on broad discovery campaigns.
• Fixed bids are useful for testing, because they remove Amazon's adjustments from the equation and let you read the effect of your own bid.
• Rule-based bidding suits established products with stable conversion data, where you can define performance guardrails and let the system stay inside them.

Change one strategy at a time and hold budgets steady while you do, or you will not know which variable moved the result. Give it at least one to two weeks before reading the outcome.

Treat Dayparting as a Test, Not a Setting

Pull performance by hour and day before changing anything. The trap is mistaking low-volume hours for inefficient hours: 3am might show a poor conversion rate on six clicks, which is not a finding. Require enough clicks in each time bucket to support a decision, change one window at a time, and re-check after a full cycle. In categories with genuine daypart patterns, this is worth real money. In categories without them, it is work that produces noise.

Do Not Cut Every Bid Equally

The across-the-board bid cut is the most popular move and the least effective, because it lowers bids on your best-converting terms and your worst ones by the same amount. Before any cut, inspect four things: search term performance, placement performance, conversion by ASIN, and cost per sale by campaign. Then reduce selectively, on the targets that are actually underperforming. An account-wide percentage is a way of avoiding the diagnosis.

The Honest Caveat

Lowering CPC is trivially easy. Bid aggressively downward and CPC may fall, but so can auction participation, premium-placement visibility, and click volume. A lower CPC is not an improvement if cost per sale or total profitable sales deteriorate. Chasing a lower click price is not the goal. Lowering ad cost per sale is, and the levers above are ordered accordingly.

Sponsored Products is the efficiency baseline and should be judged on cost per sale. It is also where most sellers' volume sits, so it deserves the most optimization attention.

Sponsored Brands costs more per click and is bought for different reasons: brand terms, defense against competitors bidding on your name, and video placements Sponsored Products cannot reach. Judging it against Sponsored Products efficiency misreads its job.

Sponsored Display is where 2026 reporting diverges most, from roughly $0.60 to about $3.72 depending on the source. Targeting type and placement drive that spread, and some Sponsored Display inventory is priced on impressions rather than clicks, which makes a blended comparison meaningless.

The practical rule: compare formats separately, always, and never judge an account on a blended CPC. A rising blended average often means your format mix changed, not that anything got worse.

Weekly, four numbers:

• CPC by ad type, never blended
• Conversion rate by ASIN, because it divides into everything
• Ad cost per sale, which is the two above combined
• TACoS, total ad spend against total sales

The fourth catches what the first three miss. Ad-attributed metrics can look stable while total business performance moves, and TACoS shows whether advertising is growing the business or recycling sales you would have made anyway.

A CMO PPC audit works the same order: CPC by ad type against category benchmarks, search term waste, placement performance, and conversion rate by ASIN. Bids are the last thing we touch, because they are the fastest lever to pull and the least likely to fix what is actually wrong.

What is a good CPC on Amazon?

There is not one in the abstract. A good CPC is one where CPC divided by your conversion rate leaves ad cost per sale below your break-even. A $1.45 click converting at 12% is better economics than a $0.72 click converting at 6%.

What is the average CPC on Amazon in 2026?

Published platform averages range from about $0.70 to $1.30, with Ad Badger reporting $1.22 in September 2026 and Epinium about $1.18. Category matters far more than the platform figure: Books averages $0.38 a click while Electronics averages $1.45.

How is Amazon CPC calculated?

You set a maximum bid, not a price. Amazon runs an auction for each eligible opportunity, where your bid and ad relevance affect whether your ad shows. Dynamic bidding, placement adjustments, and audience adjustments can all change the bid entered on your behalf, so realized CPC rarely equals your maximum.

Why is my Amazon CPC so high?

Common causes are category competition, head-term targeting, targeting breadth, bid strategy and placement adjustments stacking, seasonality, and a change in format mix. Check those before touching bids, and confirm your measurement windows have not changed what you are comparing.

How do I reduce Amazon advertising CPC?

Do not optimize CPC in isolation. Improve search term relevance, remove wasted traffic, strengthen conversion, adjust placement bids, and choose bidding strategies deliberately. Judge changes primarily by ad cost per sale rather than by whether CPC alone decreased.

Does a higher bid always mean a higher CPC?

No. A bid is a ceiling. Because you pay what was needed to win rather than your full bid, raising a bid often increases how often you win without proportionally increasing what you pay.

Do lower bids always lower my costs?

Lower bids can reduce realized CPC, but they can also reduce auction wins, visibility, and access to valuable placements. Evaluate the change against cost per sale and profitable sales rather than CPC alone.

Why is Sponsored Display CPC so much higher?

Depending on the source, it is or it is not. Reported 2026 figures range from roughly $0.60 to $3.72, because targeting type, placement, and panel composition differ enormously between providers, and some inventory is priced on impressions. Measure your own Sponsored Display separately rather than trusting any published figure.