What survives when Stripe, Shopify, and Amazon come after.

There are credibly low-to-mid millions of "power resellers" globally — people running $30k–$1M/year reselling businesses across Mercari, Poshmark, Depop, eBay, Vinted, Grailed, Whatnot, Facebook Marketplace, and TikTok Shop. They are wildly underserved. The existing tools (List Perfectly, Vendoo, OneShop, PrimeLister, Crosslist) are fragmented, pre-AI, and most are browser extensions held together with prayer.

Why your team specifically. Mercari is the only major Western marketplace whose product DNA is this seller persona. You know how the API behaves, what a power-seller's day looks like, which platforms have hostile re-listing detection (eBay) and which don't. Existing tools are built by random Shopify-adjacent founders without operator empathy. You have it.

Why it survives. Marketplaces actively don't want their sellers cross-listing — they want exclusivity. So no marketplace will build the unified layer. Shopify isn't in this market. Amazon doesn't compete in secondhand. The category is structurally hostile to consolidation by the giants.

eBay is a tired generalist. Mercari is wounded in the US. There are 30–50 verticals where used-goods commerce is large but ill-served. StockX proved sneakers, GOAT extended to luxury, Discogs owns music, Reverb owns instruments, Bring a Trailer owns enthusiast cars, Whatnot owns trading cards. The pattern: pick one category where (a) authentication matters, (b) buyers pay a premium for trust, (c) listings have rich attributes AI can extract, (d) no dominant vertical player exists.

The Mercari-relevant pick: Japanese vintage fashion sold internationally. Supply geography, language, authentication knowledge (counterfeit Japanese streetwear is a real market) — all yours. No marketplace owns this. Buyers piece together Mercari JP / Yahoo Auctions / Rakuten via proxy services that charge 10–15% on top of listing price. Customer experience is awful. A first-class English-language marketplace with AI authentication, condition standardization, and integrated international shipping could win this within 24 months.

Why it survives. Vertical marketplaces are extremely defensible once category density is achieved. eBay won't fork itself into 50 microsites. Category-specific trust signals compound. Buyers travel; sellers list where buyers are.

Organic search traffic to e-commerce sites is in measurable decline. Google AI Overview, ChatGPT shopping mode, Perplexity, Gemini, and Claude increasingly answer "what's the best ___" inline rather than sending users to merchants or review sites. Every brand is asking: how do I get cited by AI when someone asks for products in my category?

An emerging category — Profound, Goodie AI, Athena, Otterly.ai, Ahrefs Brand Radar — but early, fragmented, mostly monitoring without credible interventions. The product that doesn't exist yet is the closed loop: monitor brand citations across ChatGPT / Perplexity / Gemini / Claude / Copilot, identify why competitors are cited and you aren't, generate the structured content + schema markup + third-party citations + product feed optimization that improves your odds, re-measure.

Why it survives. Google won't build this — it's adversarial to AI Overview. OpenAI won't. Stripe won't touch it. Shopify might add a token feature, but has no LLM monitoring infrastructure.

StockX and GOAT do physical authentication for sneakers, watches, and select luxury. eBay's Authenticity Guarantee is limited and expensive. Mercari, Depop, Vinted, Poshmark, Facebook Marketplace, OfferUp — no native authentication, and counterfeit is rampant for anything over $200 AOV.

Build the white-label authentication layer any marketplace, brand, or peer-to-peer transaction can plug into. AI handles first pass (image-based fake detection for known targets — luxury bags, designer streetwear, watches, sneakers, sealed video games). High-value items route to a physical verification network — small distributed hubs in major cities, partnered experts paid per-verification.

Why it survives. Marketplaces don't want to own authentication — legal liability, operational overhead. They want to offer it as a service. So you sell to them. Brands want their own auth infrastructure for resale-program logistics. Amazon won't seriously enter because their counterfeit position is structurally compromised.

Patagonia Worn Wear, Lululemon Like New, REI Used Gear, Levi's SecondHand, Eileen Fisher Renew, Carhartt Reworked. Every major apparel and outdoor brand now operates a resale program. The infrastructure layer — Trove, Recurate, Archive — exists, but is enterprise-priced (high five-figure to six-figure ACV, multi-month implementations) and serves brands above ~$100M in revenue. The $5M–$100M brand tier is wide open.

Why your team specifically. Brand resale is marketplace-mechanics-as-a-service. Mercari is the only background that hands you the full operational toolkit — listing normalization, condition grading, fraud detection, payment escrow, dispute mediation, mobile-first UX. You aren't learning marketplaces on the job; you're productizing what you already know.

Why it survives. Shopify won't build this — their product DNA is "your own store," not "C2C inside your store." Amazon won't help brands run resale that cannibalizes new inventory and FBA. Trove and Recurate won't drop their enterprise pricing because their cost-to-serve is high — they're locked into their tier. The mid-market brand tier is structurally underserved.

Western boutique brands, independent retailers, and DTC founders perpetually want unique, hard-to-find Japanese product — premium tools, niche kitchenware, vintage textiles, indie designer fashion, specialty stationery, traditional crafts. Current options are bad: Alibaba/AliExpress (no Japan supply), Rakuten Global (badly translated, limited), trade shows (expensive, infrequent), or Japan trips (capex-heavy).

Why it survives. Alibaba doesn't do Japan supply. Amazon doesn't do upstream sourcing — they take what's offered. There is no Western competitor doing AI-native Japan sourcing. The barriers (language, business culture, network) are exactly what your team obliterates. Geographic arbitrage with an AI accelerator.