Five things from real Amazon accounts this week. All of them counterintuitive. 1
By Liranhirschkorn · May 28, 2026 · Curated by George's Blog
Five things from real Amazon accounts this week. All of them counterintuitive.
1) Sponsored Prompts can steal attribution from your own ads: A shopper clicks your Sponsored Products ad, lands on your listing, then clicks a Sponsored Prompt in Alexa shopping before buying. Amazon may attribute that sale to the Prompt. Your original ad shows a click with no conversion. The algorithm reads it as a weak keyword and pulls back spend. You paid twice for the same customer and used the second charge to penalize the first. Likey not a big issue yet but as more sponsored prompts are used on your own PDP and spend increases, these ads will need to be tested/optimized.
2) A new brand launch spreading spend across 7 ASINs at 120% TACoS isn't a campaign problem. It's a focus problem. Concentrate cash and spend on the 1–2 products with traction (or launch less products to begin with). Get those to break-even first. Trying to make everything work at once usually means nothing works.
3) A 9% TACoS can be as damaging as 60%: One brand came in well below the category average. Looked like efficiency. It was $5,000–$15,000 a month in missed sales from budget-capped campaigns and under-targeted keywords. Under-investing has a cost too.
4) Discount stacking is invisible until it isn't: Running a Best Deal, a coupon, and Brand Tailored Promotions simultaneously? They stack. One account didn't catch it until the margin damage was done. When a Best Deal or Lightning Deal is live, pause everything else.
5) Being cheaper doesn't mean winning. A STEM toy brand priced $10 below a competitor in the same category. That competitor sells 20X more units a month. Organic rank, review count, and listing quality determine the sale long before price does.