Amazon Ads CTR benchmarks: what number is normal and what a low one diagnoses

The question always shows up at the same moment: you open the Amazon Ads console, see a campaign running a 0.31% CTR, and have no idea whether that is good, bad or perfectly ordinary for what you sell. You search online, find a blog saying a good CTR is 0.5%, another saying 1.5%, a third talking about 3%, and a fourth insisting anything under 0.4% is an emergency. All four say “Amazon”, all four sound equally confident, and none of them tells you which campaigns or which country their numbers came from.

The problem is not that benchmarks do not exist; it is that a benchmark without its context is noise formatted as data. A 0.30% CTR can be excellent on a Sponsored Display placement and terrible in the first slot of a branded search. The same number, in two different situations, means opposite things. And if you sell in Mexico, there is an extra layer: almost every published statistic about Amazon Ads comes from US and European accounts, with a different competitive density, a different average ticket and a different cost per click.

This article does three things. First, it puts the published ranges on the table, saying where they come from and why they need to be handled with tongs. Second, it explains why the benchmark that matters is not the industry’s but yours, and how to build it in an afternoon. Third, it gets to the part that actually pays: when an ad falls below your own baseline, how to tell whether the problem is the creative, the price or the keyword relevance, because those are three different diseases with three different treatments.

what CTR actually measures and what it can fairly be compared to

CTR is clicks divided by impressions. If your ad showed up 8,000 times and got 28 clicks, your CTR is 28 ÷ 8,000 = 0.0035, that is 0.35%. Nothing else. It does not measure sales, spend or margin. It measures one single thing: what share of the times you appeared in front of a shopper you managed to pull them into your listing.

Precisely because it measures something so narrow, CTR only makes sense against equivalent situations. Two comparable ads need to match on at least four things: ad format, the placement where it was served, the intent behind the search that triggered it, and the product category. Change any of those four and the number stops being comparable.

The most common mistake is averaging. A seller looks at the aggregate CTR of the account, sees 0.42% and makes decisions with that. But that 0.42% is a blend of automatic campaigns landing on product pages, branded manual campaigns landing above everything, and display ads that barely scratch a click. The average describes none of the three. If you are going to compare yourself against a benchmark, the bare minimum is to split by campaign type and by placement before you even look at the number.

the published ranges by campaign type

With that warning in place, public aggregates do exist. Agencies and Amazon Ads tools compile them from their own client portfolios, and several published 2026 updates. Almost all of those aggregates come from United States accounts, not from Amazon Mexico, and that has to stay front of mind: in a marketplace with fewer competitors per search, the same creative can pull very different rates.

The ranges those compilations report by format move, in round numbers, like this: Sponsored Products between 0.35% and 0.70%, Sponsored Brands between 0.20% and 0.40%, and Sponsored Display in the 0.08% to 0.12% zone. At the whole-account level, some of these publications put the average Amazon Ads account in 2026 at roughly 0.56% CTR, with a CPC near 1.13 dollars and a conversion rate a little above 10%.

There are category differences too. The same reports place toys, art supplies and collectibles at the top, with CTR around 0.70%, followed by sports and outdoors near 0.67% and electronics near 0.62%. That ordering is what you would expect: categories where the photo sells on its own and where the shopper arrives with a clear mental picture of what they want.

Worth repeating: these are portfolio averages from US agencies. They work as an order of magnitude — they tell you that in Sponsored Products we are talking about tenths of a percentage point and not whole digits — and they do not work as a target. If your Sponsored Products campaign on Amazon Mexico runs 0.45%, you are not “below the industry average”; you are inside the range of an average that was never measured in your market.

the benchmark shifts with placement, not just with format

Here is the part almost nobody accounts for when comparing their number against a blog post. Inside a single Sponsored Products campaign, the same ad delivers wildly different CTR depending on where it was served. The 2026 compilations — again, with mostly US data — report approximate ranges of 0.5% to 0.8% for top of search, 0.2% to 0.4% for the rest of search, and 0.1% to 0.3% for product pages.

The explanation is behavioral, not algorithmic. At the top of search, the shopper has just typed what they want and is in “choose” mode. In the rest of results they are already scrolling and comparing. And on product pages your ad interrupts someone reading a competitor’s listing: your job there is not to confirm a decision but to reverse one. It is reasonable for the rate to fall by half or more.

That same placement reporting carries two figures worth keeping handy: top of search concentrates around 67% of all clicks, and its CPC typically runs 30% to 50% higher than the other placements. In other words, the placement that gives you the most clicks is also the one that costs the most, and if your placement mix shifts from one month to the next, your aggregate CTR moves even though you touched nothing.

That is why, before diagnosing a campaign, the first step is to open the placement report. An aggregate CTR that dropped from 0.48% to 0.33% may not be a performance drop at all: it may simply be that you lost share at top of search and gained cheap impressions on product pages. Same ad, same listing, worse number.

Glossary: ACoS is ad spend divided by the sales attributed to those ads; a CTR that falls while ACoS climbs almost always points to a placement or relevance shift rather than to the creative.

why the published ranges disagree with each other

Search for CTR benchmarks and you will find publications quoting 0.35%–0.70% for Sponsored Products next to others talking about 0.5%–1.2% on automatic campaigns and 1.5%–3% on manual ones. Nobody is lying: they are measuring different universes.

The differences come from four places. One, the mix of accounts: an agency serving large brands with big budgets reports different numbers than a tool used mostly by small sellers. Two, how branded campaigns are treated, since searches for your own name pull enormous CTR because the shopper was already looking for you; including or excluding them moves the average by whole points. Three, the time window, because peak season and low season look nothing alike. Four, the country: the competitive density of a mature marketplace is not the density of one still expanding.

That disagreement between sources is, by itself, the best lesson available. If four serious compilations from the same year cannot agree on what “normal” looks like, it is because the normal range depends on your case and does not exist as a universal number. The only defensible use is as an order of magnitude, while you build your own point of comparison.

how to build your own baseline in an afternoon

The good news is that the benchmark you need is already in your data. What is missing is the ordering. The procedure goes like this:

• Pull 60 to 90 days of campaign reporting, broken out by campaign, ad group and placement. Less than 60 days leaves you at the mercy of weekly noise.
• Separate branded campaigns from the rest. Searches for your own name inflate the average and tell you nothing about your ability to capture new demand.
• Group by product family, not by loose SKU. An individual SKU with few impressions has a statistically useless CTR.
• Compute the median, not the mean, inside each group. A single ad with 40 impressions and 2 clicks distorts any average.
• Store three numbers per group: median, lower quartile and upper quartile. Those three are your real baseline.

With that you have what no blog post can hand you: the normal range of your own operation, in your category, in your market. From there, the useful question stops being “is 0.35% good?” and becomes “is this ad below the lower quartile of its own group?”. That second question can actually be answered and actually leads to an action.

The baseline needs maintenance too. Recompute it every quarter and always after a large event — Hot Sale, Buen Fin, December — because those periods deform any average that includes them. A peak-season CTR compared against a February CTR is not comparing performance, it is comparing calendars.

what a low CTR diagnoses: three different diseases

When an ad falls below its own range, there are three likely causes, and they are told apart with evidence rather than intuition.

Creative. The problem is the shop window: a main image that does not read as a thumbnail, a title that does not say what the shopper searched for, a shortage of reviews against competitors with hundreds. The signature is a low CTR that stays stably low across every placement and every search term, with no event to explain it. If the CTR has been bad since launch, it is creative.

Price. Here the CTR was normal and then fell. The shopper sees your ad next to alternatives and decides on the price printed on screen. When a competitor cuts their price or a cheaper listing shows up, your rate collapses without you touching the campaign. The signature is a drop with a date: a step in the chart, not a slope. If you can date the fall and cross it against the category’s price history, the diagnosis closes on its own.

Keyword relevance. The ad is fine and the price is fine, but you are appearing on searches that are not yours. The symptom here is not just low CTR: it is low CTR with plenty of impressions and spend that does not convert. An ad for “kids school backpack” showing on “17-inch laptop backpack” will pile up impressions without clicks, and the clicks it does get will not buy either. The search term report is what resolves this case, and the fix is targeting, not photography.

The fast way to separate the three is to look at two things together: when the drop started and what happened to conversion. Low CTR forever with decent conversion is creative. CTR that falls on a date with stable conversion is price. Low CTR with high impressions and low conversion too is relevance.

the low CTR that is really an inventory problem

There is a fourth cause that does not fit the trio above because it does not live in the advertising console: your availability. When a SKU’s stock falls, the marketplace may strip your fast-shipping badge, push you down in placement or reduce your eligible impressions. The ad did not change, the price did not change, the keyword did not change, and shoppers still click on whoever promises two-day delivery.

It is the easiest diagnosis to get wrong because the data that explains it sits on another dashboard. You spend a week testing new photos while the real cause was that the SKU’s real available stock had been in the red for ten days. Before rebuilding a creative, it is worth checking the inventory for the exact period when the CTR fell: if the dates line up, you already have your answer.

Glossary: real available stock is sellable inventory net of reservations and in-transit units; when it drops you lose badge and placement, and your CTR falls without the campaign being touched.

reading CTR with its context in iqseller

The heavy lifting in all of the above is not understanding the formula, it is assembling the pieces. The CTR sits in the Amazon Ads console, MercadoLibre’s sits in Product Ads, your listing prices live somewhere else and your inventory in a third place. Rebuilding the picture of a single SKU by hand — CTR, placement, price, available stock, margin — takes an afternoon, and by the time you finish, another week of clicks has been paid at the wrong price.

In iqseller those pieces land in the same place. The Profitability module shows net margin per SKU with the COGS you loaded, the commissions from the Amazon settlement and from MercadoLibre orders, FBA and Full fees, shipping, advertising, and the VAT and withholding breakdown. The Inventory and Pricing modules put the real available stock and the price history of that same SKU right beside it, organized along the Parent → Model → SKU tree, which is the level at which a CTR can honestly be compared.

That changes the nature of the diagnosis. Instead of wondering whether 0.33% is good, you see that 0.33% next to the historical range of its family, next to the price of that period and next to the availability you had on those days. The three diseases — creative, price, relevance — stop blurring into each other because each leaves a distinct fingerprint once the data is aligned in time.

It is worth closing with a warning about the indicator itself. CTR does not pay rent: an ad can post an enviable click-through rate and still lose money on every sale if the margin cannot absorb the cost of the click. That is why a CTR baseline is built to diagnose, not to optimize in a vacuum. The number that rules at the end is still what is left after product, commissions, fees, shipping, VAT and advertising; CTR only explains why that result went up or down.