[ BDSN ] Tool finds deprecated variations before Amazon breaks your listing

💰 AMAZON RUFUS EVOLUTION: from SEARCH TO AGENTIC

BDSS 12 Virtual speaker and Amazon AI geek Andrew Bell says Amazon Rufus has crossed a critical threshold, evolving from a basic search assistant to Amazon's first true agentic commerce engine.

The biggest change is that Rufus now utilizes memory to enhance product recommendations and has developed sophisticated reasoning capabilities that fundamentally alter how customers shop.

Key New Capabilities:

Multi-Scenario Planning: Rufus can now handle multiple distinct shopping goals simultaneously. When asked to plan both a child's Star Wars party and an adult Lego party in one request, Rufus maintains separate contexts and delivers two complete, parallel plans without theme crossover.

Visible Reasoning: The AI now shows its thinking process with status messages like "thinking ..." and "gathering products ..." This transparency builds user trust by demonstrating methodical, step-by-step planning rather than mysterious black-box recommendations.

Structured Outputs: Instead of loose product lists, Rufus now provides organized blueprints with clear sections (decorations, activities, timeline, costs) that function as complete mission playbooks.

Deep-Dive Analyst Mode: Rufus can break down individual products with detailed specs, pros/cons, comparisons, and usage scenarios - essentially acting as a personal product expert.

Persona Mapping: The AI matches recommendations to specific user profiles and lifestyle needs, understanding the difference between business travelers and fashion-focused tourists.

How Rufus Works: Real Examples

Wedding Planning Orchestration

When asked to plan a backyard wedding for 50 guests with a rustic theme, Rufus created an 8-step plan covering everything from weather protection to ground anchoring.

It recommended a 29x23ft tent (noting it accommodates 100 people for extra space), coordinated seating with 50-60 chairs that matched the rustic aesthetic, and even included safety details like galvanized stakes for wind resistance.

Italy Travel Planning

For a Northern Italy trip, Rufus acted as both stylist and travel concierge, recommending a 3-piece luggage set with TSA locks for international travel, a European adapter with voltage compatibility, and clothing like a stretchy blazer "perfect for Italian restaurants, wine tours, and sightseeing" in 60-70°F weather.

Critical Shift: From Keywords to Reasoning

This represents a fundamental change from traditional keyword-based optimization. Rufus now tests real-world constraints like capacity, connectors, safety ratings, coverage, and compatibility.

Products with structured attributes and clear specifications win, while those with vague or missing data get skipped - even if they rank well for search terms.

Here is a GPT to help optimize for Rufus reasoning:

How to Prepare: 5 Action Items for Sellers

1. Provide Complete and Verifiable Data

Include structured attributes like dimensions, weight, capacity, safety ratings, and material details in standardized formats. Rufus prefers products it can reason about with confidence.

2. Optimize for Real-World Scenarios

Enrich titles, bullets, and descriptions with scenario-ready keywords. Think about specific events, themes, and use cases where your product fits (weddings, parties, travel, home projects).

3. Position Products as Complete Solutions

Create multipacks or bundles when possible. Use content and cross-referencing to highlight how your products complement others in customer missions. Products that "play nice" in groups get recommended more often.

4. Leverage Reviews and Customer Q&A

Analyst Mode draws heavily on reviews and Q&A content. Encourage detailed customer feedback about performance, durability, and use cases. Populate Q&A with specific scenario answers.

5. Design for Multi-Scenario Flexibility

Products serving multiple roles across different plans gain more visibility. String lights can work for weddings, parties, or home décor - highlight these varied use cases so Rufus includes your product in more plan types.

Rufus marks the shift to agentic commerce where AI doesn't just recommend products - it plans and executes complete shopping missions. Success now depends on making products easy for AI to reason about, context-ready for missions, and indispensable to curated bundles.

The competitive advantage goes to sellers who position their products as solution components within broader customer missions, not just standalone items. In this new era, reasoning is the quality gate, and compatibility with AI-driven plans determines market success.

Read Andrew’s full deep-dive here.

🕹️ URGENT: AMAZON VARIATION THEME PURGE on SEPT 2

Amazon is eliminating deprecated variation themes starting on September 2, 2025 affecting over 65,000 themes. If you don't fix affected listings, Amazon will automatically split your child ASINs into standalone products, causing you to lose:

• Aggregated reviews

• Combined product rankings

• Traffic concentration

• Ad spend efficiency

Immediate Action Required:

1. Audit Now: Download Amazon's removal list or new product templates (deprecated themes marked "Do Not Use")

2. Fix Affected Families:

   • Delete old parent ASIN

   • Create new parent with approved theme

   • Reattach child products

   • Verify front-end display and review aggregation

3. Prioritize Strategically: Start with top sellers, but don't force inappropriate themes just to maintain families

Critical Warnings:

• Even basic themes like Size, Color, Style, and Flavor are being removed in some categories

• Variation misuse can trigger account suspension

• Changes may break listings as early as September 2nd

• Q4 timing makes this especially disruptive

Seller Concerns: Many worry that deleting/recreating parents could trigger variation abuse flags, causing weeks of lost sales. The inconsistent removal criteria (useful themes cut while odd combos remain) adds confusion.

Vanessa Hung of Online Seller Solutions created a nifty Amazon Theme Checker that makes this entire process much easier.

Check it out here

🙉 SHOPIFY’S DIRTY LITTLE SECRET WINS SMART CONTEST

This strategy from Matt Kostan won 1st🏆 place at the Billion Dollar Sellers Summit 12 S.M.A.R.T. contest last Thursday. You'll see Shopify's dirty little secret (and the quick code tweak) that keeps AI recommending your brand instead of your competitors. 

The full replays from BDSS 12 Virtual come out this week - snag yours today.

📈 PATTERN, AMAZON’S #1 SELLER FILES to GO PUBLIC

Pattern started in 2013 when co-founders David Wright and Melanie Alder launched "iServe" from Melanie's living room in Lehi, Utah, with $50,000 each.

They began by reselling refrigerator magnets and other small products on Amazon. Melanie ran the business largely by herself for the first two years while David kept his day job, working on the company in the evenings.

The founders refused to take a paycheck for nearly three years and famously had to ask Melanie to reinvest her first paycheck just days after receiving it to fund new inventory.

Pattern Group, now a major Amazon marketplace accelerator, filed for its IPO on August 22nd, seeking to raise approximately $400 million. Here are the details Amazon sellers need to know from their S-1 filing and recent coverage:

Impressive Financial Growth: Pattern posted $47M net income on $1.14B revenue in the first half of 2025, representing a 35% revenue jump compared to $35M net income on $841M revenue in the same period last year. For full-year 2024, they achieved $67.9M net income on $1.8B in revenues.

Strong Operational Metrics: Pattern maintains a 4.5x inventory turnover rate and 96% in-stock rate in North America, demonstrating the kind of operational excellence that Amazon sellers strive for.

Proven Brand Partnership Model: Nearly half (48%) of revenues come from brand partners working with Pattern for 5+ years.

Health & Wellness Dominance: 63% of Pattern's inventory purchases focus on health and wellness brands, with their two largest brands representing 28% of total inventory purchases.

Heavy Amazon Reliance: Pattern's revenue heavily depends on Amazon (93.6% of total revenue, with 87.9% from Amazon specifically), only 5.3% from other marketplaces like Walmart, Target, eBay, and TikTok Shop, and just 1.1% from SaaS services.

AI and Data Focus: The company positions itself as using "AI-driven tools that optimize sales for brands across global marketplaces" emphasizing data analytics capabilities throughout their filing.

IPO Timeline: Working with Goldman Sachs and JPMorgan, Pattern expects to launch their roadshow after Labor Day when the IPO market typically reactivates.

From selling magnets in a living room to a $2B valuation, Pattern's IPO validates the massive opportunity in professional Amazon marketplace management.

For sellers, it underscores that with the right data-driven approach and relentless focus on operational excellence, even small Amazon sellers can build billion-dollar businesses.

Curious to know more - here is their IPO S-1 document with details

📛  CHAT GPT-5 COOKBOOK HELPS ECOM WORKFLOWS

OpenAI has released its official GPT-5 Prompting Guide, revealing proven techniques from their internal training that could revolutionize how e-commerce businesses use AI.

The standout technique uses structured, step-by-step collaboration: "You are a methodical AI assistant. Here's the task: [insert task]. Provide a step-by-step plan first, then execute only the first step upon approval. Continue after confirmation."

• Better Control: No more AI running wild with assumptions on product descriptions, marketing copy, or inventory planning

• Iterative Collaboration: Plan → confirm → execute approach perfect for complex tasks like campaign development or customer service protocols

Practical Example: Instead of asking for a complete product listing, try: "You are a senior e-commerce strategist. Outline steps to optimize this product page for conversions. Then implement only the headline optimization first."

The Big Shift: Moving from simple "give me an answer" requests to structured, collaborative workflows. This approach is particularly valuable for complex e-commerce tasks like multi-channel marketing campaigns, inventory optimization, and customer journey mapping.

✍🏽 STRUGGLING to GET PRODUCT REVIEWS?

GetReviews has helped 100's of Amazon sellers worldwide collect real and honest reviews by designing review funnels and generating unique QR codes for package inserts.

Add these inserts to your packaging to get customers to leave genuine reviews. Also offer giveaways such as warranties, promo codes, e-gift cards or digital products.

What's the win for you here?

• compatible with Shopify, TikTok Shop, Walmart and more!

• set up to be compliant with Amazon TOS

Schedule a call today to learn more.

Don't wait - improve your review strategy today.

Use code BDS15 to get 15% off your first 3 months.

✌🏼 See you again Thursday …

The answer to today’s STUMP BEZOS is
58% of Prime Members purchased groceries on Walmart