Amazon Says Rufus Saves Shoppers 20%. That Number Should Worry You.

Amazon put a number on its official page about Rufus: customers using Auto Buy save an average of 20% per purchase.

Amazon means this as a selling point for the feature. For brand managers running Spring Sale discounts in four days, it should read as a warning.

That 20% doesn't come from nowhere. It comes from shoppers setting price targets inside Rufus, waiting months if necessary, and completing purchases when products hit their threshold. Every one of those successful purchases was a full-price sale that didn't happen — replaced by a discounted one, on the shopper's timeline.

The question isn't whether Auto Buy is good for shoppers. It clearly is. The question is what it means for brands that run aggressive sale events, and what those sale events write into Rufus's memory for the months after.

How Auto Buy Actually Works

A shopper finds your product at $34.99. They want it but aren't ready to commit at that price. They tap the Auto Buy option in Rufus and set a target: "Buy when it drops to $27 or less." Rufus monitors the price for up to six months. When the price hits $27, Rufus completes the purchase using saved payment and shipping information — no cart, no checkout page, no second decision from the shopper.

This is the documented feature. Not a speculation. Prime members have used it since mid-2025.

Separately: Rufus analyzes 90-day price history on products when making recommendations. If it detects aggressive price swings, it can tell shoppers to wait for a better price. This is also documented — Amazon confirmed the behavior, and sellers started noticing it in Rufus responses last year.

Put these two mechanics together. A shopper sees your product at $34.99 and either sets an Auto Buy target or asks Rufus "is now a good time to buy?" Rufus consults the 90-day window. If the 90-day window shows your product recently traded at $24.99 — which it will, after the Spring Sale — the recommendation may be to wait.

The difference from seeing a price history chart on Google Shopping is that Rufus surfaces this proactively. The shopper doesn't have to look for it. Rufus volunteers it when they ask a price question — or sometimes when they don't. It's part of the response, not a tool the shopper has to remember to check.

Amazon is advertising that Auto Buy users average 20% savings per purchase because the system is actually delivering that. The sale prices that produce those savings are coming from your catalog.

What Seven Days of Discounting Does to the Rest of Q2

Amazon's Big Spring Sale runs March 25–31. If you drop your hero ASIN 25% for that week, Rufus has that data point for roughly 10 weeks — the entire 90-day window. Every shopper who encounters your product between late March and early June is interacting with Rufus that knows your product recently traded at 25% off.

The practical risk breaks into two scenarios.

Scenario one: passive shoppers. A shopper sees your product, hasn't decided yet, and asks Rufus "is this a good deal?" Rufus has context from the 90-day window. Depending on how it weights recent price data (Amazon hasn't published its exact methodology), the response may include a mention that this item went on sale recently. The shopper waits.

Scenario two: Auto Buy queue. A shopper who saw your product at full price before the sale may have set a price target. If your Spring Sale hits that target, Rufus completes the purchase at the sale price. That's a conversion that was going to happen anyway — just at your margin, not yours.

Neither scenario means the Spring Sale is wrong. But it means the margin calculus extends past March 31.

Where This Is Worst

Supplements and regulated categories are the most exposed. Amazon Rufus already strips efficacy claims from these listings — so price signals and review sentiment carry disproportionate weight in Rufus responses. A supplement brand that runs $12.99 during sales and normally sits at $19.99 is writing a price signal Rufus will reference for months. The supplements comparisons show how aggressively these categories discount — which is exactly the pattern that creates the strongest pricing anchors.

Baby products are sensitive because Auto Buy makes intuitive sense for repeat purchases. Parents buy diapers and formula on schedules. A parent who sets an Auto Buy target after seeing a price drop during the Spring Sale has established what they expect to pay — potentially for the next six months. The baby products guide covers Rufus's heavy query volume in this category. Those queries increasingly involve price context.

Health monitors and cameras attract shoppers who explicitly ask Rufus price comparison questions ("is this a good price for a blood glucose monitor?", "what did the Sony ZV-1F sell for last month?"). For these categories, 90-day price history isn't background context — it's a direct input to the query. The health monitors comparisons and cameras guide capture how much price intelligence shoppers expect in these Rufus conversations.

Subscription-eligible products compound this. If a shopper auto-subscribes after a sale discount, that becomes their reference price for every reorder — not the MSRP.

What To Do About It Before March 25

You have four days and probably a discounting plan already in place. This isn't about scrapping that plan. It's about being deliberate inside it.

Know your floor and hold it. If your profitable discount threshold is 15%, don't let automated repricing push the sale to 22% because competition is aggressive. If you're using a repricing tool (Feedvisor, Seller Snap, Wiser), check whether you've set a minimum floor — most have this setting, it's just rarely configured with Rufus price history in mind. Set it before March 25. The incremental volume at 22% likely doesn't offset writing a lower price floor into Rufus's 90-day window. Set the minimum and lock it manually for the sale period.

Be selective about which ASINs you discount. Don't touch your catalog anchor — the product shoppers use to calibrate what your brand costs. That's the one Rufus references when making relative price comparisons. If you discount your flagship 30% for a week, you've told Rufus (and every shopper who uses it) that your flagship goes for 70% of its listed price.

Check Auto Buy eligibility on your high-volume ASINs. Open the Amazon shopping app and search your products. If the Rufus interface shows price monitoring or Auto Buy options, your pricing history on those ASINs is actively being used by shoppers as a trigger. Know what targets are likely set before you run the price down.

Think in quarters, not weeks. The Spring Sale revenue is real. So is the price signal. A brand that drives $200K in Spring Sale volume at 28% off and then loses 15% of its April and May conversion rate to "wait" recommendations from Rufus is doing math it didn't intend to do.

The Measurement Gap

None of this is visible in Amazon's native reporting.

The Ads Console doesn't show you when Rufus told a shopper to wait. You can't see how many Auto Buy targets your Spring Sale triggers. Amazon doesn't report "impressions where Rufus cited price history" or "sessions that ended with a 'check back later' recommendation." The data that would let you measure this mechanism doesn't exist in Seller Central.

What you can do: track what Amazon Rufus actually says about your products and category across the sale period and the weeks after. Does Rufus mention recent pricing in its responses? Does it recommend waiting in your category? Does your visibility in Rufus responses drop post-sale relative to competitors who held price?

That monitoring layer is what AgentBuy provides — tracking what Rufus says about your brand and products across query types, including price-related commentary and how your visibility in Rufus responses shifts over time.

The Spring Sale is worth doing. Just go in knowing that it sets more than this week's revenue. It sets Rufus's expectations for the next quarter.