Rory Wingo. a college sophomore and daughter of ChannelAdvisor and Retailgentic founder Scot Wingo wrote a guest post for her dad's Retailgentic newsletter describing how she and her friends shop.

Her generation doesn't even say "ChatGPT" anymore. It's just "Chat." Skincare routine for dorm hard water? Ask Chat. What to wear for sorority recruitment? Ask Chat. Worth $30? Ask Chat.

The key insight: TikTok still owns discovery. AI now owns everything after.

Shopping has always been a conversation. You tried on the dress, stepped out of the fitting room, and asked your mom or your friends: "does this look good on me?" You sent screenshots to the group chat: "pick one."

That conversation isn't going away, but it's moving to AI. The difference is Chat remembers every previous conversation: that she wears gold jewelry, likes neutrals, and will pay more for something that lasts. Her friends need the backstory every time. Chat doesn't.

Rory's own example: she saw a viral lip combo on TikTok.

The old journey was pause the video → dig through comments → Google → Sephora → TikTok Shop → compare → buy. Instead she tapped TikTok's built-in AI assistant and asked "what lip combo is this?" It identified every product and shade in seconds. "Where can I buy them?" — retailers, prices, availability. "Is there a TikTok Shop option?" — listings, bundles, ratings, tutorials.

The funnel compressed from:

→ TikTok → Google → Reddit → Sephora → Amazon → Buy to: → TikTok → AI conversation → Purchase

Notice what's NOT in that second funnel.

What this means for you:

→ Discovery and validation have split. Creators and TikTok spark the want; AI answers "should I actually buy this?", "which shade for me?", "where's it cheapest?", and "is there a bundle?" You now have to win both moments, and the second one has no ad unit.

→ You're no longer just competing for clicks. You're competing for AI recommendations. If the assistant answers "is this worth it?", your product data, reviews, comparisons, and content need to give AI the information and trust to confidently recommend you. Thin listings and vague claims won't get surfaced in a conversation.

→ The platforms are pushing this on purpose. Rory was served an Instagram ad from ChatGPT itself — not for a product, but for the habit of asking AI what colors suit you before you shop. The recommendation is becoming part of the product.

→ This is agentic commerce from the shopper's side. Today AI helps decide. Tomorrow it compares retailers, builds the cart, applies the coupons, and checks out. For Gen Z, that won't feel like a revolution, just the next message in a conversation they're already having.

Your youngest customers aren't asking their friends if the dress looks good anymore. They're asking Chat. Make sure Chat has something good to say about you.

Credit: Rory Wingo, guest post on Scot Wingo's Retailgentic. Worth reading in full.

If you are still obsessing over getting backlinks to rank higher on Google, you are playing a game that ended in 2015.

In this episode of the Marketing Misfits, hosts Norm Farrar and Kevin King sit down with Jason Barnard, the "Brand SERP Guy" and founder of Kalicube, to discuss how AI engines like ChatGPT, Gemini, and Perplexity are completely changing the way consumers discover and buy products.

Jason explains why you need to stop thinking of AI as a search engine and start treating it as an "untrained sales force". If you do not actively educate these algorithms on who you are, what you do, and why you are the best, they will simply recommend your competitors.

Whether you run a local brick-and-mortar store or a massive e-commerce brand, Jason breaks down exactly how to control your narrative across Google's Knowledge Graph, structure your product pages for machine readability, and use AI as an orchestrator for your content.

🔥 Join the Marketing Misfits Newsletter: https://misfits.news

Andrew Bell spotted Alexa for Shopping's latest A/B test: AI Overviews appearing in a desktop sidebar on broad category searches. He thinks this one sticks.

Here's what's inside the module:

→ A category insight synthesized from sources on AND off Amazon — sometimes personalized. In his rug example, Alexa referenced the household's kids and multiple pets (never typed into the search) and reordered the pathways accordingly: Machine Washable and Pet-Friendly before Wool and Plush Shag. Same query, different guidance per shopper. Your Alexa visibility may depend on the shopper's inferred circumstances, not just the keyword.

→ Visual navigation that writes the next search for you. Click "Nature" on a metal wall art search and Alexa drops "nature metal wall art" (or "nature-inspired metal wall art") straight into the A9 search bar. Alexa is doing the query reformulation shoppers used to do by retyping — which means longtail coverage in your listings matters more, not less.

→ Prompt pills trained on clicks. The "Dive Deeper" prompts that appear are the ones shoppers click most — the marketplace is learning which research questions matter per category, in public.

→ External sources feed the overview. The earbuds example cites SoundGuys, PCMag, and What Hi-Fi — editorial content sitting upstream of the Amazon shelf. Bell's advice: work on getting into those cited sources (he calls it "authority anchoring"). Off-Amazon PR is now on-Amazon SEO.

On mobile, this already runs inside the results feed as "Researched by Alexa" — sources, insight, prompts, and scrollable visual pathways, all before the shopper ever reaches sponsored and organic listings.

Two grounding notes from Bell worth repeating: COSMO is a knowledge graph, not an algorithm — it did NOT replace A9, whatever LinkedIn tells you. And Amazon SEO is alive and well; Alexa is just using the search bar more intelligently than shoppers do. Oh, and Walmart is copying all of it.

The Wall Street Journal's Patrick Coffee reports that AI-generated product demos, makeup applications, and dress twirls are spreading across TikTok Shop. The tension between creators, brands, and TikTok itself is boiling over.

The setup: TikTok Shop itself launched an AI Video Maker that lets affiliates auto-generate shoppable videos. Platform policy allows fully AI-generated content to earn commissions, as long as it carries an on-screen "AI-generated" tag and makes no misleading claims.

Creators are turning still images into 3D videos where their avatars appear to hold and sample products. One creator told the WSJ she pumps out at least one fully AI promo post per day and teaches others to do the same: "I can create whoever you need me to.”

The backlash is real:

→ SharkNinja fired the first big shot — a memo to its affiliate community banning AI-generated content outright: use it and your commissions get removed. "We didn't want an AI-generated Shark vacuum cleaning an AI-generated floor," said CCO Neil Shah.

→ Rare Beauty got AI-twinned without consent. A creator posted videos of her AI duplicate using Rare Beauty products. All commission-eligible sales, zero connection to the brand. Because Rare Beauty is in TikTok's open affiliate plan, they couldn't stop it.

→ Human affiliates are steamed — real-product reviewers say AI videos are getting the ad spend and taking the sales.

TikTok Shop's US GMV is projected to hit $23.41B this year (+48%), passing Target and Costco online. Affiliate ranks exploded from 2.3M creators in 2024 to 11.3M globally — 945K in the US.

If your brand is on Shop, the control levers are limited but real: switch from the open affiliate plan to targeted (you hand-pick creators), tighten creator approval, and disable the setting that auto-converts top-performing posts into paid ads.

There's no way to fully stop AI posts made "on your behalf," so at minimum, someone on your team should be reviewing how your products are showing up. Your brand may already be starring in videos you've never seen.

Amazon has finished rolling out Search Query Performance (SQP) and Search Catalog Performance dashboards to Vendor Central . This has been a query-level data that's been a 3P-only advantage since it launched.

It's not a perfect match to what sellers get (Amazon is hiding the market-share view from vendors), but the core is there:

→ Real search volume & rank for the exact terms driving buyers to your products
→ Full-funnel visibility by search term and ASIN: → clicks → add-to-carts → purchases
→ Brand or ASIN view — see data aggregated at the brand level or drilled down to individual ASINs

Why this is a big deal for vendors:

→ PDP optimization with actual data. You can now see which queries you're winning impressions on but losing clicks or cart adds. That's your title, image, and price problem laid bare.

→ Smarter Sponsored Products. Steer ad spend toward the queries where the funnel shows you convert, instead of guessing from campaign-level data.

Where to find it: In Vendor Central, hover over Brands → Brand Analytics → Search Analytics → Search Query Performance (or Search Catalog Performance).

If you're a vendor who's been flying blind on organic search while your 3P competitors mined this data for years, your homework just got assigned.

✌🏼 Have a great weekend.

See you again on Monday.

The answer to today’s STUMP BEZOS is
51% of all AI-researched purchases close on Amazon