Several products in one Amazon ad group: Amazon picks who gets the clicks
A Sponsored Products ad group can hold several products. It is a common setup: you put the whole product line or all sizes of a product in one ad group, give it the keywords, and let it run. The keywords or targets and their bids belong to the ad group, so every product in it competes with the same bid. Which product shows up for a search is decided by Amazon, not by you.
We checked how that decision turned out on real accounts. In most ad groups one product took most of the clicks. And that product was often not the one that converted best.
The measured numbers below come from the Sponsored Products advertised product report of 12 advertiser accounts in Europe, the UK and the US, day by day from March 28 to October 7, 2026. Not every account ran for the whole period. We stopped a week before the last available day so that the ad-attributed sales are close to final. That is 36,960 daily rows and 63,592 clicks. Every comparison is made inside one ad group, so currencies never mix.
How common it is
Of 608 ad groups, 118 showed at least two products, 488 showed one, and 2 showed none in the period.
How much money runs through these ad groups depends on how the account was built. In the euro accounts 29.7% of the Sponsored Products spend went to ad groups with several products, in the pound accounts 55.3%. The dollar accounts had one product per ad group, so there the share was zero.
One product takes most of the clicks
We kept the ad groups with several products and at least 100 clicks: 31 ad groups in 6 accounts. The median ad group had 2 products, the largest one 16.
- The most clicked product took a median 71% of the ad group's clicks.
- In 14 of the 31 ad groups it took 80% or more.
- Impressions were split the same way: the most shown product got a median 73% of the impressions.
So "the ad group advertises the whole line" is mostly not what happens. Mostly one product is advertised, and the others get what is left.
The product with the most clicks is often not the best one
To compare conversion rates (orders divided by clicks), a product needs enough clicks. 22 ad groups had at least two products with 20 or more clicks each. In these:
- The most clicked product had the highest conversion rate in 7.
- In 7 it had the lowest ACOS.
- In the other 15, another product of the same ad group converted better.
Many of these differences were small. In 7 of the 15, the better product converted less than 30% better, and on these click counts that can be chance. In 4 the difference was large enough that chance is unlikely (a standard test for two rates, at 95%).
The result holds with stricter limits. With at least 50 clicks per product, the most clicked product was the best converting one in 6 of 16 ad groups. With at least 100 clicks, in 4 of 10.
An example from one euro account. Two products in one ad group, with the same targeting and the same bids:
- Product A: 795 clicks, 83 orders, 10.4% conversion rate, 0.68 euros per click, 45.7% ACOS.
- Product B: 180 clicks, 31 orders, 17.2% conversion rate, 0.73 euros per click, 29.0% ACOS.
Product A got 82% of the ad group's clicks and 537 euros of spend. Product B got 132 euros. The product that sold at a much lower ACOS got the smaller part of the budget, and no bid inside that ad group could change this, because the bid belongs to the ad group, not to the product.
A second example from a pound account, an ad group with several products. The most clicked product had 184 clicks, 12 orders and 14.4% ACOS. Another product had 77 clicks, 11 orders and 8.1% ACOS, at the same 0.30 pounds per click. It brought almost as many orders from less than half the clicks.
This is not a test of why Amazon chose this way. The report shows how the clicks were split, not the reason. What it shows is that the split did not follow conversion in most ad groups.
What we could not see
The report practically only lists products that were shown: of the 339 product rows in ad groups with several products, only 1 had zero impressions. So a product that Amazon never showed at all is invisible here, and the real concentration can be even higher. The report counts purchases for 7 days after the click, and an order can be for another product of yours than the one in the ad. Inside one product line, where buyers switch sizes or colors, this blurs which product really converted. We did not have margins, so "best" here means conversion rate and ACOS, not profit. The 100 click and 20 click limits are our choice, not a measurement.
Check it in your account in 20 minutes
- Download the Sponsored Products advertised product report for the last 60 days, ending at least a week ago.
- Make a pivot table by campaign, ad group and advertised ASIN with impressions, clicks, spend, orders and sales.
- Keep the ad groups that have more than one ASIN. For each, calculate what share of the clicks the top ASIN took.
- Add conversion rate (orders divided by clicks) and ACOS for each ASIN that has at least 50 clicks.
- Look for ad groups where an ASIN with fewer clicks has a clearly better conversion rate or ACOS than the top ASIN.
- For such a product, decide whether it deserves its own bid. If yes, move it to its own ad group with the same keywords or targets and pause its ad in the shared group, so that the two do not compete with each other.
If you put several products in one ad group, you are letting Amazon decide which of them gets the budget. Check once a month whether it chose the one you would have chosen.