Skip to main content
Where an ad shows changes what its click is worth more than almost anything else: a click at the top of the search results turns into an order about 1.4 times as often as a Rest of Search click on the same keyword (1.8 times pooled across keywords), yet costs only 10 to 25% more. Top of Search is underbid on the accounts we manage, and Detail Page clicks are overpriced.

The three placements

A placement is the spot on Amazon where an ad is shown. Amazon reports every Sponsored Products click and order under one of three placements:
  • Top of Search (ToS): the sponsored slots at the top of the first page of search results.
  • Rest of Search (RoS): every other sponsored slot in search results, lower on the first page and on later pages.
  • Detail Page (DP): ads on other products’ detail pages, the page a shopper lands on after clicking a product.
One keyword bid competes for all three. An advertiser can raise the bid for one placement with a placement modifier: a percentage added to the base bid, set once per campaign, up to +900%. The auction explains what a bid buys. We work on the belief that Top of Search and Rest of Search are filled by one auction per search, while Detail Page ads come from a separate auction. We have not measured this.

How the placements differ

Three rates describe a placement:
  • Click-through rate (CTR): clicks divided by impressions, where an impression is one display of the ad.
  • Conversion rate (CVR): orders divided by clicks, counting orders Amazon attributes to the click within 7 days.
  • Cost per click (CPC): what a click actually cost.
Accounts we manage, US Sponsored Products, non-branded keyword rows, 90 days, pulled September 2026: Read down the CVR column: a Top of Search click converts 1.8 times as often as a Rest of Search click and 2.4 times as often as a Detail Page click, for about 10% more per click. Pooled numbers mix different keywords, since the keywords that win Top of Search are not a random sample. Comparing the same keyword across placements (keywords with at least 100 clicks in both) gives a smaller but still large gap: Top of Search converts at a median 1.39 times Rest of Search, with the middle half of keywords between 1.12 and 1.79 times. Its click-through rate is a median 6.8 times higher and its click price 1.24 times higher. Across whole accounts (those with 500 or more clicks in each placement) the conversion ratio is a median 1.82.

The Top of Search price premium depends on the price level

On cheap searches a Top of Search click costs much more than a Rest of Search click; on expensive searches it can cost less. Grouping keywords by their Rest of Search click price:
“Top of Search is the expensive placement” holds only for cheap searches. Above about $3 a click, Top of Search costs the same or less per click than Rest of Search, while still converting better.

What a Top of Search modifier buys

A Top of Search modifier shifts an advertiser’s impressions toward the top of the page. Across our accounts, the share of an advertiser’s own impressions that land at Top of Search rises with the modifier: Top of Search carries 43 to 53% of an advertiser’s clicks even though it holds a small share of its impressions, because its click-through rate is so much higher. The same lever, seen from the search’s side, on one account (60 generic searches, August 2026): the share of all searches on which its ad won Top of Search went from 0.5 to 1% with no modifier, to 2 to 8% at +1 to +99%, about 15% at +100 to +199%, and 40 to 80% at +200% and up.

Where money is won and lost

In July 2026 we measured how profit responds to bids using randomized bid changes. Three findings:
  • Top of Search is underbid. Each extra dollar of Top of Search bid returned 11 to 15 cents of profit, so those accounts would earn more by bidding higher there.
  • Value per click follows conversion. The estimated value of a click was $7.19 at Top of Search, $3.69 at Rest of Search and $3.52 at Detail Page, while click prices across the three are within about 10% of each other.
  • Detail Page clicks are overpriced. We checked whether late-arriving orders explain the gap; they do not (attribution timing explains none of it on the two accounts measured).
Conversion also varies over time: by up to 2.1 times across the hours of the day, and to 0.92 times its weekday level on weekends.

What we have measured