Two policy changes rolling out right now will gut the "Was $X / Now $Y" badges that have quietly propped up conversion rates across the Amazon marketplace. If you've ever parked a List Price in Seller Central that your product has never actually sold for, Amazon is about to catch it.

Here's what's changing and what you need to do before it hits.

Change #1: List Price verification (live today, April 23, 2026)

The strikethrough number next to your price now has to be real. Amazon will only accept a List Price if the product has recently sold at that price through another reputable retailer, or buyers actually purchased it as the Featured Offer on Amazon at that price.

If neither is true, the strikethrough disappears. No more picking a fantasy MSRP to inflate the apparent discount.

Change #2: Typical Price recalculation (live May 18, 2026)

The Typical Price is the median paid over 90 days, and right now it ignores promotional sales. That's how sellers run nonstop coupons while still flashing a fat "save 30%" badge against an untouched baseline.

On May 18, that loophole closes. If your price sits below the non-promotional median for more than half of any 90-day window, Amazon folds those promo prices into the baseline. Your Typical Price drops. Your strikethrough shrinks. Your savings badge may vanish entirely.

Why this hurts

The strikethrough and the "you save X%" callout are two of the strongest conversion triggers on a PDP. Sellers who have been manufacturing the perception of a deal are about to watch conversion rates slip without anyone touching their ad spend or creative.

There's also legal pressure behind this. A proposed class action in Washington state accused Amazon of leaning on fictional list prices to sell Prime Day. Regulators in multiple regions are tightening rules on advertised savings across ecommerce. Amazon wants its discounts to look credible heading into Prime Day 2026.

What to do this week

The bigger picture

This is Amazon choosing long-term shopper trust over short-term deal theater. Sellers who built real price discipline will look sharper. Sellers who depended on optical discounts are about to find out how much of their conversion was the badge, not the product.

Check your List Prices today. The first deadline is already here.

Out of the $700 billion spent on digital marketing in 2024, a staggering $140 billion was stolen by fraudsters? If you are running digital ads, there is a high probability that 20% to 25% of your traffic is entirely fake, generated by bots, malware, and human fraud farms.

In this eye-opening episode of Marketing Misfits, Rich pulls back the curtain on the dark side of Pay-Per-Click (PPC) advertising, sharing insane stories of competitors using cheap bot software to drain rivals' ad budgets by up to $15,000 a day.

Discover why blindly trusting big tech platforms is costing you money, how to block fake clicks, and the immediate settings you need to change in your advertising dashboard to protect your ROI.

Click here to join the AI Amazon PPC Challenge now

^ Learn to use Claude Cowork and other agentic AI tools to take over whole roles inside your Amazon brand ^

Get a library of plug and play Claude skills and copy-and-paste prompts ready to use right away.

See the exact AI agents and workflows 8 figure sellers are currently using inside their brands.

Most sellers building AI agents fail the same two ways. They stall after one mediocre attempt and decide agents aren't ready. Or they sprawl into twenty disconnected tools that produce more maintenance than value.

The missing piece is architectural. You need an answer to "which agents does my business need, and how do they connect?" Here's the five-role framework that works for ecom operations.

The Five Roles

1. Triage Agent — The Front Door Sorts and routes every inbound signal. Pre-sale questions, post-purchase issues, returns, wholesale inquiries, supplier emails. Categorizes, tags, and logs everything in structured format.

Build this first. It's the lowest-risk agent (it routes, doesn't respond), produces structured data that fuels every other agent, and gives you back 60-90 minutes a day immediately.

2. Intelligence Agent — The Analyst Pulls from your Triage data and combines it with external research. Product review monitoring, competitor pricing, seasonal demand signals, audience language shifts.

The leverage: when customer inquiries about sizing spike at the same time a competitor launches a sizing tool, you see converging signals, not two disconnected data points. Produces weekly briefs you scan in 10 minutes.

3. Draft Agent — The Writer First drafts of product descriptions, email campaigns, social content, supplier communications, listing copy.

This is the agent most sellers want to build first. Don't. Without Triage and Intelligence upstream, it writes from training data and produces generic output. With them, it writes product page copy that addresses the actual objections customers raise in pre-sale tickets. Big difference.

4. QA Agent — The Gatekeeper Reviews everything before it leaves the building. Listing accuracy, compliance checks, brand consistency across channels, factual accuracy, voice match.

Most sellers skip this and become the bottleneck themselves. QA exists so you can scale Draft Agent volume without quality dropping. Flags issues, sends back to Draft with revision notes, escalates only the exceptions to you.

5. Operations Agent — The Dashboard Monitors the other four agents plus your business metrics. Inventory velocity, return rates, campaign performance, channel profitability, plus each agent's output volume, quality scores, and error rates.

Builds last because it needs the data the first four create by running. Delivers a 5-minute daily summary instead of you checking every system.

Why the Order Matters

Each agent makes the next one better. Triage creates structured data. Intelligence uses it to produce grounded briefs. Draft uses those briefs to write in your customers' actual language. QA scores the drafts. Operations monitors the whole system.

Build them in isolation, you get sprawl. Build them in sequence, you get compounding.

The Audit (If You Already Have Agents Running)

Map every agent currently running in your business to one of the five roles. Score each role:

Most sellers find the same pattern: multiple agents in the Drafting role, nothing in Triage, nothing in Operations. All throttle, no steering. Fix it by building backward. Stand up Triage. Within 30 days the categorized data improves your existing Draft agents without touching a single prompt. Then add Intelligence, QA, and Operations.

What to Do This Weekend

Pick one. If you're starting from zero, scope your Triage Agent. If you already have agents running, run the audit. Either way, the output is a one-page spec per role covering what it does, what data it needs, and where its authority ends. That's the document that turns a stalled or sprawling AI stack into a system that compounds.

Five roles. One backbone. Build in order.

The problem most sellers miss: You launched a new and improved version. The old listing is still live, still pulling traffic, still holding reviews. But when shoppers land there, they hit a dead end, or worse, a competitor.

BDSS Dream 100 member Vanessa Hung recently broke down a case where a supplement brand had 44 ASINs stuck in this exact gap. Old liquid drops, new capsule versions, zero connection between them. Here's how she closed it.

Variations ≠ Newer Version

Sellers confuse these constantly.

Variations are for the same product in different sizes or colors. Same detail page, switchable options.

Newer Version is a separate relationship type for genuine upgrades. Same function, same intent, improved format. Both ASINs stay as separate listings. Amazon adds a "Newer version available" module on the old detail page that points shoppers to the current offer.

Trying to force an upgrade into a variation = policy risk. Use the right tool.

The Submission: Two Fields, 24 Hours

No catalog restructuring. No parent-child hierarchy. No rebuild.

Alternate path: you can also submit through Seller Assistant chat inside Seller Central by giving it both ASINs.

Why This Matters

Sellers using the Newer Version Widget cut lost sales during stock transitions by 20-30%. Click-through from the old listing to the new one runs 15-25%.

That is demand you already paid to generate. Reviews you already earned. Rank you already built. Without the link, it evaporates every time the old ASIN goes out of stock or a shopper lands there first.

Quick Audit Questions

The Bigger Lesson

Amazon's catalog has relationship types that are not surfaced in the main Seller Central menu. Not visible does not mean not available. It means you have to go looking.

The system does not assume a connection between two ASINs just because the connection is obvious to you. Every relationship has to be submitted explicitly. If you assume Amazon "gets it," you are leaving the connection unbuilt and the traffic stranded.

Any time a customer's path through your catalog depends on a link between two ASINs, test whether that link has actually been submitted. Most of the time, it has not.

Based on Case #018 from Vanessa Hung, Online Seller Solutions. Her team handles Newer Version submissions for sellers who do not want to deal with Seller Central directly.

✌🏼 Have a great weekend.

See you again on Monday.

The answer to today’s STUMP BEZOS is
Meta is expected to do $243B in ad revenue in 2026