Amazon Just Started Charging for Rufus Ads — What Sponsored Prompts Mean for Your PPC

On March 25, 2026, Amazon flipped a switch most advertisers haven't fully reckoned with. Sponsored Products Prompts and Sponsored Brands Prompts — the AI-powered ad enhancements that surface inside Rufus conversations and on shopping result pages — exited open beta and became billable. Every Sponsored Products and Sponsored Brands campaign in the US was auto-enrolled. There was no opt-in screen, no migration plan, no grace period. If you're running Sponsored Ads on Amazon today, you are already running Rufus Ads. The question is whether you're optimizing for them or paying for impressions you can't read.

What Sponsored Prompts Actually Are

Sponsored Prompts are AI-generated, contextual product surfaces that appear inside the Rufus shopping assistant and at moments-of-decision across product detail pages and search. When a shopper asks Rufus a conversational query — 'What's the best magnesium for sleep if I'm sensitive to citrate?' — Rufus draws from your campaigns and surfaces a small set of products as the answer. Click into one, and the conversation continues inside Rufus or expands on the page where the prompt appeared.

Crucially, this is not keyword targeting. The shopper never typed 'magnesium glycinate.' They asked a question. Amazon's AI decided which products belonged in the answer based on your structured backend data, recent conversion velocity, return rate, recency of reviews, and category authority. Your bid is one input among many — and it is no longer the dominant one.

The 50-to-5 Discovery Collapse

The most consequential shift is what Rufus does to the result count. A traditional Amazon search returns 48–60 organic results plus sponsored placements across multiple pages. Rufus, in contrast, typically surfaces five to seven products as a direct answer to a conversational query. Industry analyses through Q1 2026 estimate Rufus is now mediating between 15% and 20% of mobile shopping queries.

Compress the funnel from 50 results to 5, route 1 in 5 mobile queries through that compressed funnel, and the implications stack quickly:

• Long-tail keyword strategies that depended on ranking somewhere in the top 30 organic results lose most of their conversion path on Rufus-mediated traffic.
• Position 1–3 organic dominance becomes proportionally more valuable than ever, because positions 4–50 may no longer exist for that query.
• Brands without strong category authority — measured in reviews, recency, return rate, and Subscribe & Save penetration — are systematically excluded from Rufus's shortlist before any auction even runs.
• PPC stops being a way to buy a click. It becomes a way to buy candidacy in the AI's answer set.

Why Structured Backend Data Now Outweighs Ad Copy

Multiple practitioner analyses through 2026 converge on the same finding: structured backend attributes — not lifestyle imagery, not premium A+ content, not clever ad copy — are the primary lever determining which products Rufus surfaces. Rufus reads the structured fields first because they are the cleanest signal of what the product actually is.

For supplement brands specifically, the backend fields that most reliably influence Rufus inclusion are:

• Item type keywords and category nodes — accuracy here determines which conversational intents Rufus considers you for.
• Ingredient attributes filled at the structured level (not just in bullets) — Rufus parses these, not paragraphs.
• Dosage form, count, serving size, and dosage strength as discrete attributes.
• Certifications as enumerated values: NSF, USP, Informed Sport, organic, third-party tested.
• Backend search terms — still relevant, but now treated as a synonym map for Rufus, not as a keyword stuffing field.
• Subscribe & Save enrollment status, which functions as a category-authority signal Rufus weights heavily for repeat-purchase categories.

The brands winning the most Rufus impression share in early 2026 are not the ones with the best ad copy. They are the ones whose product detail pages are structured cleanly enough that an AI can answer questions from them without hallucinating.

The Shadow Penalty for Returns and Recent Sentiment

Rufus does not just look at star rating. It weights recent return rate and recent review sentiment as gating factors. A product with a 4.6-star average over five years but a 22% return rate over the last 90 days is materially less likely to appear in Rufus's shortlist than a 4.4-star product with an 8% return rate. Industry observers have called this a shadow penalty: it is invisible in your traditional reporting because the impressions you lose were never logged as auctions you entered.

What Supplement Brands Specifically Need to Do in the Next 30 Days

1. 1Audit every backend attribute on your top 10 ASINs against the structured field schema for your category. Empty fields are missing answers. Any attribute Amazon offers as a structured input should be populated.
2. 2Pull a 90-day return rate report by ASIN and flag anything above 12%. For flagged ASINs, pause aggressive PPC scaling until the return driver is identified — you are paying for impressions Rufus is suppressing anyway.
3. 3Read your last 30 days of reviews for each hero ASIN and tag the recurring complaints. Address them in product copy, customer service responses, and — where applicable — formulation or packaging changes. Rufus reads recent sentiment, not just stars.
4. 4Map your top 20 keywords to the conversational queries a real shopper would ask. 'Magnesium glycinate 400mg' becomes 'best magnesium for sleep without citrate side effects.' Rewrite bullet points and A+ content so the literal answers to those questions appear in your listing.
5. 5Test Sponsored Brands Prompts deliberately. Even though enrollment is automatic, the campaigns you let Amazon enhance versus the ones you actively shape produce different impression mixes. Build a dedicated SB campaign for your top three ASINs with custom creative that reads like an answer, not an ad.
6. 6Increase exact-match defense on your branded terms. Rufus is more likely to surface competitors as alternatives than legacy search ever was. The cost of letting a competitor's Sponsored Prompt appear inside a Rufus answer to your branded query is substantially higher than the cost of defending the keyword.

Budget and Bid Implications

Through Q1 2026, accounts that started actively optimizing for Sponsored Prompts saw two consistent patterns: total impression count rose modestly, but the impression mix shifted toward higher-intent placements. Click-through rate on Prompts placements has run materially above traditional search-result placements in early reporting. Conversion rate has been mixed — strong for brands with clean backend data, weak for brands relying on copy and imagery to do the persuasion work.

The right budget response is not 'spend more.' It is 'reallocate.' Cut budget on broad-match campaigns that were generating volume against generic queries. Move that budget into exact-match defense, brand-protection Sponsored Brands campaigns, and a controlled Sponsored Display retargeting layer that recaptures shoppers who saw your product in a Rufus answer but didn't convert in-conversation.