Amazon said on 30 July 2026 that Amazon Ads generated $19.8 billion in revenue in the second quarter of 2026, up 26% year over year. The figure appeared in Amazon's own newsroom post, published alongside its Q2 2026 results, in which CEO Andy Jassy said that "Sponsored Products continues to be our largest offering and a key driver of growth."
Read the rest of this as Amazon's own marketing
Every performance figure in Amazon's 30 July 2026 post is self-reported by Amazon about its own advertising products, in a corporate newsroom post promoting them. Amazon published no methodology, no sample size, no time window and no independent verification for any of the percentages below.
That does not make the numbers wrong. It makes them a vendor's claims, which is a different category of evidence from a measured result, and worth holding at that distance when a media plan is being justified with them.
Amazon says Sponsored Prompts shoppers convert 48% more often
Amazon said on 30 July 2026 that shoppers who click Sponsored Prompts convert 48% more often and spend 21% more than shoppers who do not. The format appears on Alexa+ and Alexa for Shopping, which places it inside conversational shopping rather than on a results page.
The comparison is the thing to look at. Shoppers who click an ad inside an assistant conversation are already further along than shoppers who do not click anything, so a gap between the two groups is expected before any credit goes to the format itself. Amazon did not publish how it controlled for that.
Amazon says its Ads Agent cut cost per impression by 8%
Amazon described Ads Agent, an AI campaign tool it says cuts campaign setup from hours to minutes, and reported 8% lower cost per impression and 6% lower cost per acquisition. Amazon said the tool expanded to 11 new countries over the year, and did not name them.
Setup time is the claim with the least ambiguity here, because it is a workflow change rather than an outcome. The cost figures are averages across an unstated set of advertisers.
Amazon says its sports inventory sold out in Q2 2026
Amazon reported that its streaming sports inventory sold out across Thursday Night Football, NBA, WNBA and NASCAR, and that over 30 new advertisers joined NBA coverage in the property's first year. It also said brands advertising across multiple sports see 2.3x higher unduplicated reach than single-sport advertisers.
Amazon added that multi-sport viewers drive 12% higher spending and 17% more orders on Amazon, a claim that connects its ad inventory to its retail business. Sold-out inventory is the more checkable of the two statements, because it is a fact about supply rather than an effect on behaviour.
The numbers, and who is making the claim
The table below separates the figure from its source, because in this story they are the same party for every row.
| Claim | Figure | Who says it |
|---|---|---|
| Amazon Ads revenue, Q2 2026 | $19.8bn, up 26% year over year | Amazon, reported with Q2 2026 results |
| Sponsored Prompts vs shoppers who do not click | +48% conversion, +21% spend | Amazon, self-reported |
| Ads Agent campaign performance | 8% lower cost per impression, 6% lower cost per acquisition | Amazon, self-reported |
| Multi-sport vs single-sport advertisers | 2.3x higher unduplicated reach | Amazon, self-reported |
| Multi-sport viewers on Amazon | 12% higher spending, 17% more orders | Amazon, self-reported |
Only the revenue line carries the weight of a reported financial result. The rest are product claims published to sell the products they describe.
What this means for Thai marketers
Amazon's post does not mention Thailand and does not list the markets where Sponsored Prompts, Ads Agent or the streaming sports formats are available, so anything about local relevance is reasoning rather than something Amazon said.
For most Thai brands, Amazon Ads is an export play rather than a domestic channel: it matters if you sell cross-border into markets where Amazon retail operates, and it is close to irrelevant if your demand sits on domestic marketplaces and social commerce. A brand already running cross-border ecommerce into the US, the UK or Japan has a reason to read the Sponsored Prompts claim closely. A brand selling only in Thailand does not.
The transferable part is not the ad product at all. It is that shopping queries are moving into assistant conversations, which is the same shift already visible in how people search. Product detail that is clear, complete and machine-readable is what gets surfaced in that format, which is a product page and feed quality problem before it is a media-buying one. Treating it as content work rather than a budget line is the cheaper starting point, and it is not gated behind a market launch.
Frequently asked questions
Is Sponsored Prompts available in Thailand?
Amazon's post does not say. It named Alexa+ and Alexa for Shopping as the surfaces, but published no list of markets and did not mention Thailand at any point.
How much did Amazon Ads make in Q2 2026?
Amazon Ads generated $19.8 billion in Q2 2026 revenue, up 26% year over year, according to Amazon's 30 July 2026 post published with its Q2 2026 results.
Are the 48% conversion and 21% spend figures independently verified?
No. Both are self-reported by Amazon about its own ad format, published in a newsroom post promoting it, with no methodology, sample size or time window disclosed.
Which 11 countries did Ads Agent expand to?
Amazon did not name them. The post says the tool expanded to 11 new countries over the year without listing which ones, so there is no way to tell from the source whether any are in Southeast Asia.
Where to start
If cross-border marketplaces are already part of your revenue, the useful question is whether your product data is good enough to be surfaced by an assistant, not whether to buy a new ad format. Relevant Audience can look at that with you.
