What is CTR?

CTR (click-through rate) is the share of your ad's impressions that turned into clicks. It measures one moment in the shopper's path: they saw your ad among the other results and chose it.

Definition

Amazon Ads describes CTR as the "percentage rate of shoppers who click on your ad after seeing it" (Amazon Ads, measuring campaigns, checked 2026-10-03). On Amazon the ad is usually a product tile among other results, so CTR mostly reflects what that tile shows next to its neighbours: the main image, title, price, star rating and delivery promise.

On a cost-per-click campaign an impression nobody clicks costs nothing. CTR is therefore not a profit metric on its own. It decides how much traffic a given amount of visibility produces, and it is the first place to look when visibility is not turning into visits.

Formula

CTR = Clicks ÷ Impressions × 100

When you roll CTR up across keywords, placements or days, divide total clicks by total impressions. Never average the individual rates.

Worked example

Example inputs, invented for round arithmetic: one keyword over 30 days, split by placement.

The keyword's real CTR is 1.0%: 200 clicks from 20,000 impressions. Averaging the two placement rates gives 1.75%, which overstates it by three quarters, because product pages carried four times the impressions.

The same table shows how a blended CTR can fall when nothing about the ad changed. Suppose next month product pages take 26,000 impressions at the same 0.5%, which is 130 clicks, while top of search holds at 4,000 impressions and 120 clicks. Combined CTR drops to 250 ÷ 30,000 = 0.83%. Every placement performed exactly as before; only the mix moved.

Carry it one step further. At a 1.0% CTR, 20,000 impressions give 200 clicks, and at an 8% conversion rate those clicks give 16 orders. CTR and conversion rate multiply, so a weakness in either one caps the other.

How to use CTR to make decisions

Use it to find where the funnel breaks. A shopper goes from impression to click to order, and CTR covers the first step. Low CTR with healthy conversion points at the search result tile. High CTR with weak conversion points the other way: the tile promises something the detail page or the price does not deliver, or the search is broader than the product.

Suspect relevance first. A keyword with plenty of impressions and very few clicks is often matching searches your product does not answer. Check the search term report before rewriting the listing. A negative keyword may fix it outright.

Compare like with like. CTR means something against the same keyword, placement and ad type over time, or against the market for the same search. Brand searches, where the shopper asked for you by name, behave so differently from generic ones that they need to be read separately.

Measure yourself against the market on the same query. In the Search Query Performance report, your CTR relative to the market's equals your click share divided by your impression share. Example inputs: if you hold 5% of a query's impressions and 4% of its clicks, your CTR is 0.8 times the market's on that search. Shoppers see you and pick someone else, and the fix sits in the tile, not the bid. That report counts organic and paid traffic together, so read it as a listing signal rather than an ad one.

Common mistakes

1. Borrowing someone else's "good CTR". Generic CTR tables mix categories, ad types, placements and brand traffic, so a figure from one tells you nothing about another. For an outside reference point, Amazon Ads benchmarks reporting compares your CTR and CPC with category peer brands in the console and the API. It became generally available worldwide on 18 May 2026 for brands registered in Brand Registry (Amazon Ads announcement, checked 2026-10-03).
2. Averaging rates. As the worked example shows, the mean of several CTRs is not the CTR of their total.
3. Reading tiny samples. Three clicks from 200 impressions is a 1.5% CTR. One more click would make it 2.0%. Wait until a single click no longer changes the conclusion.
4. Chasing CTR at the expense of the sale. An image or price that wins clicks the page cannot convert raises CTR and spend together. On a cost-per-click campaign, what pays is clicks that turn into orders.
5. Assuming a higher CTR buys cheaper clicks. Amazon says its auction weighs factors beyond the bid when it sets placement and final CPC, and its CPC guide does not list them. Treat any promise that better CTR lowers CPC as unproven.
6. Mixing filtered and unfiltered counts. Amazon filters invalid traffic, and for sponsored ads its gross and invalid traffic report shows gross clicks and impressions beside the invalid ones it removed (Amazon Ads, checked 2026-10-03). Compare gross with gross and filtered with filtered.

How CTR relates to adjacent metrics

Clicks = Impressions × CTR
Orders = Impressions × CTR × CVR
CPM    = CPC × CTR × 1,000

Impressions set the ceiling on clicks and CTR converts them. Conversion rate then turns clicks into orders. On a cost-per-click campaign CTR shapes volume rather than price. On a campaign billed per thousand impressions it shapes price as well, because the same impression cost is spread over more or fewer clicks. Example inputs: at a $5 cost per thousand impressions, a 0.5% CTR makes each click cost $1.00 and a 1.0% CTR makes it cost $0.50.

How Shurq helps

• CTR on every row. The /ad-manager grid shows CTR beside impressions and clicks from campaigns down to search terms, with the change against a comparison period when you set one.
• Word-level CTR. N-gram analysis in /ad-manager breaks your search terms into single words and phrases with CTR, conversion rate and CPC for each, so a word that draws impressions but no clicks or orders stands out, and can go straight to the negatives flow.
• You against the market. The Search Query Performance funnel in Shurq sets your click-through rate beside the market's on each query, and flags queries where your biggest gap is at click-through: shoppers see you in the results but click a competitor.
• Exposure kept apart from performance. Shurq's bid optimizer treats a target that is barely being shown as an exposure problem and gives it a deliberate nudge upward, unless its campaign is already running over target.