English NewsSunday, 15 November 2026
Latest
Amazon

Seller briefing: billing, reference prices, AI ads and marketplace visibility

2026-04-06

Illustrative miniature shopping cart, not a seller dashboard or a study chart.

Illustrative miniature shopping cart, not a seller dashboard or a study chart.

An April 6, 2026 seller briefing argued that billing, fulfilment charges, reference pricing and AI advertising were changing the economics of online retail at the same time. Kevin King described the April and May changes as pressures on working capital and margins, while reporting early Rufus advertising data and a study of Walmart's visibility in Google Shopping. This is a dated account of the supplied briefing and its examples, not confirmation of current platform terms or an account-specific instruction.

Advertising payments and the working-capital squeeze

The briefing said that from April 15 Amazon would deduct advertising costs from seller proceeds, using the saved card only when proceeds were insufficient, with no opt-out described. It argued that this would remove card rewards on ordinary ad payments. At a stated 2% reward rate, its examples were $12,000 a year lost on $50,000 monthly ad spend, $24,000 on $100,000 and $60,000 on $250,000.

King's estimate used Amazon's reported $68.6 billion advertising revenue for 2025, up 22%, and assumed 30% was card-paid at about 2.5% interchange. That produced roughly $20.6 billion in card volume and $515 million in potential annual fee savings. Those percentages were the briefing's assumptions, not disclosed payment-mix measurements.

The note linked the change to DD+7, described as holding proceeds for seven days after confirmed delivery before disbursement eligibility. It said sellers had previously combined a roughly 30-day card-payment window with about 30 days of held proceeds to manage an approximately 60-day cash-flow buffer. Steven Pope characterised the combination as a double hit to liquidity.

The briefing compared Google's move away from cards for high-spending advertisers from mid-2024 with Meta notices in February 2026 requiring affected accounts to move to monthly invoicing by March 31. It also described a one-time $2,500 Amazon promotional advertising credit for affected accounts. King considered that small beside the ongoing loss of rewards and rising bids; it was not presented as cash or a universal payment to every seller.

A further announced pressure was a 3.5% fuel and logistics surcharge on FBA fulfilment fees from April 17, averaging about 17 cents per unit but varying with size and dimensions. The briefing said Amazon linked it to higher fuel and logistics costs during the Iran war. Sellers quoted in the note doubted whether it would disappear. Together with the April 23 reference-price change, King described three pressures within eight days.

Why frequent discounts could reset the comparison price

For April 23, the briefing described new List Price validation: a genuine purchase at that Featured Offer price on Amazon, or evidence of that price at another retailer. For May 18, it described a Typical Price calculation that would include promotional sales if the Featured Offer stayed below the non-promotional median for more than half of the previous 90 days. A habitual discount could therefore become the comparison baseline rather than a visible saving.

Its example was a $40 product with a 20% coupon for 60 of 90 days. The note said a roughly $32 Typical Price could leave the same $32 selling price without a crossed-out $40 comparison. Unlabelled price reductions were described as non-promotional sales. Subscribe & Save, tailored coupons, Buy X Get Y and peak-event sales such as Prime Day were listed as exclusions at that time.

King argued that consistent real-world pricing across a brand's own shop and retail partners would help validate genuine comparisons, while always-on coupons could lose their visual advantage. He warned that stopping discounts on May 18 would not instantly restore the old baseline: around 90 days at or above the non-promotional median could be needed under the calculation he described.

Rufus prompts: little traffic, useful information

The April briefing described sponsored prompts as paid suggested questions in search results and product pages that opened a branded Rufus conversation. It said charging was per click. The early sample was tiny: under 1% of Sponsored Products clicks, with one seller making $20 million to $30 million annually receiving 88 prompt clicks against about 500,000 other ad clicks since January.

The briefing reported roughly $0.31 per click compared with $0.50 to $0.70 for conventional Amazon ads, and higher click-through when prompts appeared despite low volume. It cited 140% year-on-year Rufus usage growth and more than 300 million users over a year. The reported immediate value was seeing which prompts triggered ads and which products appeared, then using that information to improve listing and storefront explanations, including why one version cost more.

Limitations mattered. Many prompts already included the brand name, advertisers could not see the exact answer the shopper received, and generated questions could alarm buyers. One advertiser reportedly opted out of a prompt asking about a chemical that was not in its product. The note said existing advertisers were automatically enrolled and urged account review, while treating the format as an experiment rather than a major sales driver.

Marketplace seller names and Google Shopping visibility

A Productrise study cited in the briefing examined more than one million listings across over 50,000 Google Shopping result pages in the first quarter of 2026. It identified more than 1,600 Walmart seller names, such as marketplace names prefixed with Walmart. The claimed mechanism was that Google treated displayed seller identities separately even when the links ended at Walmart.com, supported by individual Merchant Center sub-accounts.

The note described carousels as commonly having eight slots and reported perfect seller diversity in only 13.7% of them. Regrouping Walmart identities into one entity reduced perfect diversity by 32% in carousels containing Walmart, according to the study. One example had eight Walmart seller names across three carousels on a page. These were study findings as relayed in the briefing, not evidence that each seller was fictitious or a finding of a rule breach.

King argued that marketplaces could occupy more search space as they recruited sellers, an advantage unavailable to a single-name independent retailer. The note acknowledged real third-party merchants behind the names and described divided expert reactions. It connected the strategy with Walmart's reported retreat from OpenAI instant checkout after weaker conversion than its own website, while putting more weight on traffic to its own store.

Figures, commercial notices and brief items

An included TikTok Shop graphic stated 97% year-on-year sales growth from brands above $30 million, 103 billion US e-commerce searches in 2025, 80% growth in total transaction volume and 146% growth in creators earning commissions, with more than 16,000 generating six-figure sales. A separate CreatorIQ chart showed the top 10% receiving 53%, 60% and 62% of annual creator payment volume in 2023, 2024 and 2025; the top 1% received 15%, 18% and 21%. The two panels describe different measures and should not be treated as one dataset.

A historical Nashville Ecom Mastery promotion offered live streaming, on-demand access and replays, saying access was limited to 200 people. A separate Stack Influence promotion described product-only micro-influencer payments, outside traffic, reusable image and video content rights and automated collaboration management. It named Magic Spoon, Unilever and MaryRuth Organics and cited a Blueland case after Shark Tank, claiming up to 13-fold monthly revenue growth in two months and a separate 13-fold return measure. These were promotional claims, not interchangeable metrics or promised results. A 10% sign-up discount belonged to that month's offer.

The short reading list pointed to a critique of TACoS as a metric, Amazon and Walmart competing beyond the purchase button, the fuel surcharge and an Amazon-FedEx returns expansion described as adding 10,000 drop-off locations. Those were linked headlines, not full additional articles supplied in the email. A delivery quiz introduced the claimed end of the Amazon-USPS delivery partnership, said Amazon represented 15% of USPS shipments and gave 1.7 billion annual deliveries as its answer. The closing thought urged readers not to let delay defeat their goals. The shopping-cart photograph is illustrative and does not reproduce any seller dashboard or study chart.

More Amazon business news

More in Amazon business