On Amazon, Kevin King described a new AI system underneath the platform that learns not just what shoppers search but WHY they buy, using purchase signals, reviews, and behavioral data.

What is the name of that semantic understanding engine?

Answer at bottom of email

Kevin and Norm had spent years living and breathing Amazon keywords. Helium 10 open on one monitor, Jungle Scout on the other, backend keywords stuffed to the character limit.

It worked. Until it didn't.

The shift wasn't dramatic. It was a webinar Kevin and Norm hosted, walking sellers through something Amazon had quietly released: Cosmo.

Not the chatty AI assistant everyone knew.

Something deeper.

A semantic understanding engine living underneath the platform, learning not just what shoppers searched, but why they bought.

Kevin showed an example on screen: a shopper searches "shoes for pregnant women" and clicks on a pair of slip-resistant sneakers. No keyword match on pregnancy anywhere in the listing.

But Cosmo had learned from millions of purchase signals that pregnant women buying footwear prioritize safety, not style. The algorithm made the connection that no copywriter had written explicitly.

The room went quiet…

Then came Rufus, Amazon's AI shopping chatbot sitting on top of Cosmo. By Prime Day 2025, roughly 50% of younger shoppers had interacted with Rufus at least once. They weren't typing keywords.

They were asking questions.

"What's the best protein powder for endurance running?"

"What should I get a new mom who doesn't sleep?"

The old playbook of repeating target keywords seven times across bullets and titles was suddenly competing with conversational AI that pulled from reviews, Q&A, product attributes, and behavioral data all at once.

The marketing lesson is blunt: Amazon is no longer a keyword search engine. It's an intent engine.

Sellers who adapt listings to speak to buyer psychology will win. Those who keep counting keyword density are already falling behind.

Amazon quietly went through one of the most significant platform shifts in its history. Two AI systems, Cosmo and Rufus, have fundamentally changed how products get discovered, compared, and purchased. Kevin King's breakdown of these systems revealed the exact adaptation playbook sellers need to survive.

Ritu Java, CEO of PPC Ninja, added critical context by connecting Amazon's shift to the Five Levels of Awareness framework, a Eugene Schwartz classic that's more relevant today than when it was written.

The common thread across both episodes: keyword-centric Amazon marketing is becoming obsolete.

Intent, semantic relevance, and behavioral signals now drive rankings. Sellers who get this, and adapt their listings, creatives, and ad strategy, will pull away from the pack.

Watch These Misfits Episodes to Dive Deeper:

The Decoy Effect is one of the most powerful and underused pricing psychology strategies in e-commerce and marketing. Here's the premise: introduce a third pricing option strategically positioned to make one of the other two look like the obvious best value, and buyers predictably choose that "best value" option at higher rates than they would without the decoy.

The classic example: a movie theater sells a small popcorn for $3, a large for $7, and a medium for $6.50. The medium is the decoy. It makes the large look like almost the same price for dramatically more product. Large sales spike.

How to apply this in your business:

Step 1: Identify your target offer. Which product, tier, or SKU do you most want customers to buy? That's your "target." Everything else gets designed to make it look compelling.

Step 2: Build your decoy. Price it close to but slightly less than your target, with noticeably less value. It shouldn't be a good deal. It's there to make the target look great by comparison.

Step 3: Set your anchor. Your premium tier should be priced high enough to make the target feel like a bargain, but not so expensive it creates sticker shock. Most buyers will anchor on the premium price and pick the target as the "smart" middle ground.

Step 4: Test on Amazon. Use Amazon's variation listing to run three tiers of the same product (different sizes, bundles, or quantities). Watch which tier captures the most conversions. Adjust your decoy pricing until the middle tier dominates.

Step 5: Apply to SaaS, services, and info products. This isn't just for physical goods. If you sell courses, memberships, or retainer services, a three-tier pricing structure with a decoy mid-tier dramatically increases average transaction value.

Real-world example: Norm Farrar's knife brand (discussed in MMP #009) introduced a $224 premium tier specifically to make the $124 mid-tier feel like a steal. The $224 tier rarely sold, but it increased $124 tier conversions by measurable margins.

Amazon quietly went through one of the most significant platform shifts in its history. Two AI systems, Cosmo and Rufus, have fundamentally changed how products get discovered, compared, and purch

Each AI platform has distinct content preferences and citation patterns. Google AI Overviews pull from its live index, so your SEO foundation must be strong. Perplexity values authoritative content with clear citations, and draws heavily from Reddit, YouTube, and review platforms, making multi-channel presence important.

ChatGPT favors comprehensive, conversational content. Its training data skews toward major publications and widely-cited sources, making digital PR and third-party citations especially valuable. Microsoft Copilot combines GPT-4 with Bing's live index, making Bing rankings matter.

The unified AEO approach: build content that's structured, authoritative, and answer-first (this works across all platforms), then add platform-specific signals. Reddit community presence for Perplexity, strong Bing rankings for Copilot, major publication citations for ChatGPT, and Google E-E-A-T for AI Overviews.

For help with email marketing for your brand, contact dragon.fish

The Brand: Hims & Hers launched in 2017 with a simple but audacious mission: make embarrassing healthcare conversations easy. Hair loss, erectile dysfunction, skincare, mental health.

Topics men and women avoided at the doctor's office became the center-piece of a bold DTC brand built on humor, transparency, and radical accessibility.

The Strategy: Hims attacked stigma head-on. Their early ads featured direct, un-clinical language and playful creative that made men feel completely normal about problems they'd never discuss openly.

"Tackle hair loss before it tackles you" wasn't just clever copy. It was permission structure. By making the conversation feel casual and modern, Hims attracted customers who would never walk into a pharmacy and ask for these products.

They paired this brand positioning with a DTC subscription model, a telehealth platform connecting customers to licensed providers, and a pricing strategy that dramatically undercut the traditional healthcare system. A $20/month hair loss subscription removed every barrier.

The Result: Hims went public via SPAC in 2021 and has grown to over $1 billion in annual revenue as of 2024. Their playbook proved that healthcare stigma isn't a market limitation.

It's a marketing opportunity. Any brand in a "taboo" or uncomfortable category can apply the same logic: name the discomfort, make it normal with humor and transparency, and remove friction from the purchase.

The Lesson: Your most defensible competitive advantage might be the conversation your competitors are too afraid to start.

The "figurado" cigar shapes (torpedo, pyramid, perfecto) are the hardest to roll and require a master roller with 10 or more years of experience. In some factories, only 2 to 3 rollers out of 200 or more are even qualified to attempt them.

Every figurado is shaped entirely by hand, tapered from both ends or pointed at the head, demanding a level of muscle memory that takes a decade to develop.

It's the reason figurados command premium prices and why counterfeit versions are almost always visually off.

The next Collective Minds Society Cigar & Whisky trip is Feb 18-22, 2027

Amazon's AI systems reward intent-based content over keyword density. Here's how to rebuild your listings for the new reality.

Sample Prompt:

"Act as an Amazon shopper. Ask 10 natural-language questions someone would ask Rufus when looking for [my product]. For each question, identify whether my current listing copy directly answers it."

The seller who writes for buyer intent, not keyword count, wins the Rufus recommendation.

What is Amazon Rufus and how does it affect sellers?

Amazon Rufus is an AI-powered shopping assistant launched in early 2024. It responds to conversational, intent-based queries from shoppers rather than traditional keyword searches.

For sellers, Rufus changes how products get discovered: listings written in natural language that directly addresses buyer intent are surfaced more often than keyword-stuffed listings.

What is Amazon Cosmo and how is it different from Rufus?

Cosmo is Amazon's semantic understanding engine operating underneath the platform. It learns from purchase behavior, click patterns, and reviews to understand the intent behind searches.

Rufus is the visible chatbot interface shoppers interact with. Cosmo is the AI that powers Rufus's understanding of why shoppers buy, not just what they search.

How should I change my Amazon listing to rank better with AI?

Rewrite product bullets using Feature plus Benefit pairings that address buyer psychology directly. Add a Q&A section using natural-language questions your buyers actually ask.

Use lifestyle images that show who uses your product and in what situation. Remove redundant keyword repetition and replace it with conversational, intent-rich copy.

Do keywords still matter on Amazon in 2026?

Yes, but they matter differently. Keyword relevance still signals category, but repeating exact-match keywords multiple times no longer boosts ranking the way it once did.

Cosmo rewards semantic relevance: does your listing speak to the intent behind the search? Sellers who combine strategic keywords with conversational, benefit-led copy outperform pure keyword-stuffed listings.

Cosmo.

In the Marketing Misfits webinar, they revealed that Amazon's Cosmo is the semantic understanding engine running beneath the platform. It learns from purchase signals, behavioral data, and reviews to understand the intent behind searches, not just the words.

Rufus, the visible AI shopping assistant, sits on top of Cosmo and uses its intelligence to answer conversational shopper queries.

Watch the full conversation