As marketplaces expand offline, they will face regulation (e.g. Uber, Airbnb). Ideally, leading marketplaces set their own regulatory standards within categories.

Yes — many marketplaces are using AI / automation: chatbots, FAQ bots, auto-routing tickets, auto-drafting replies, etc. These reduce human workload while maintaining or improving service speed.

Have a clear 15–20 slide deck; target the right stage/sector VCs; prefer warm intros; concise emails with credentials, traction (GMV/take, margins), raise size; reasonable round sizes.

Stage and geography agnostic; roughly ~55% US/Canada, ~25% Western Europe, ~10% Brazil/India, ~10% rest of world (incl. Eastern Europe, Africa, SE Asia).

Companies like Hero Stuff, Rebag: you take a few photos, record a video/story, transcribe voice to keywords, then listing is generated (category, title, description, price). Also generate short snippet videos for social.

Moving from listings (Craigslist, Gumtree) to transactional ‘instant buyers’ (fair, leasing models) that let sellers get money the same day, improving the experience.

Shift from listings/classifieds (Craigslist, Gumtree) to transactional and leasing models (Fair, Auto1). Instant-buyer models allow sellers to get money quickly with better UX.

Investing in logistics, payments, robotization, warehousing, export / import services. Examples: Flexport, ShipBob, Stripe, Mundi (SMB exporters fintech doing collections, FX, cargo insurance, etc.).

Applied AI for listing and productivity. Examples: Rebag’s “Claire”.

Marketplace-pick model: instead of RFQ and user sorting, the marketplace selects the supplier (like Uber picks the driver) using availability, price, terms, etc.

Old model: users posted needs and browsed suppliers (Thumbtack). New model: integrated verticals (Block Renovation, Sutter) complete the work for you, reducing friction.

Early token rewards (especially when network value is low) can help bootstrap supply and/or demand, reduce marketing spend, and improve unit economics.

They often begin with paid marketing to attract users, but over time, word-of-mouth reduces reliance on paid acquisition.

They can allow marketplaces to outsource high-impact work to the community (unsalaried contributors), scale virtual teams, and lower fixed costs, enabling lower take-rates.

Technology is deflationary and marketplaces make things cheaper, giving more people access and leveling opportunity. They also enable efficiency gains that reduce waste and emissions.

For suppliers: handle logistics, payments, and operational tasks. For demand: offer a better user experience, often prioritizing convenience over price.

They assess four things: (1) Team – visionary, eloquent, and able to execute; (2) Business – TAM and unit economics; (3) Deal terms – fair for stage/traction/category; (4) Thesis fit – aligns with their focus areas (inequality of opportunity, climate, mental/physical well‑being).

High local concentration (duopoly/monopoly) removes pricing power, pushing you toward brokerage; fragmentation at the local level is key.

Redefining the listing process: take photos + a short description/video → generate full listing, category, price, title. Also used in customer care and programmer productivity. Examples: Hero Stuff, Rebag.

Within 10–15 years, cooking at home may become a luxury. Cloud kitchens, autonomous delivery, and economies of scale will make delivered meals cheaper and higher quality than cooking at home.

The concern is that LLMs could capture the “top of funnel” discovery phase: people might start product searches there, or even make purchases through them, bypassing marketplaces.

Keep take rates reasonable (~10–15%) and provide real value (payments, insurance, escrow, trust, convenience). Favor fragmented, non-monogamous relationships (e.g., rides) where disintermediation is less attractive.

Yes — if you already index in Google, you should also index in LLMs (i.e. make your content discoverable by AI answer engines) to benefit from that traffic. The exception is when a marketplace has ~95% market share in a vertical: in that rare case, indexing might train users to start their journey with LLMs / search engines rather than the marketplace itself.

Failing to find product-market fit (can’t get CAC/LTV to work); co-founder issues; raising too much at too high a price. Marketplace-specific: onboarding too much supply early—most marketplaces are demand-constrained.

Pre-seed: ~$1M at ~$5M pre. Seed: ~$3M at ~$10–12M pre, with ~$25–30k/mo net revenue (e.g., $150k GMV at 15% take). Series A: ~$7M at ~$23M pre for ~$150k/mo net revenue. Series B: ~$15M at ~$50M pre for ~$500k/mo net revenue.

Several: low margins; price sensitivity; operational intensity; switching incumbent behaviors; need to pick industry wisely; concentrated supply or demand; difficulty monetizing in categories with thin margins.

Cross-border commerce; Live commerce; New goods verticals supported by service layers; High take rate business models based on convenience; Climate tech; AI-enabled marketplaces; B2B marketplaces (marketplace for inputs; SMB enablement; Friend-shoring; B2B labor marketplaces; Business infrastructure; B2B recommerce).

High average order value (AOV) or high frequency, or both; fragmented supply and demand; markets where buyers/sellers are willing/able to shift from legacy methods; enough margin and willingness to pay a take-rate in the future.

1. B2B Marketplaces for Inputs 2. SMB Enablement 3. Friendshoring 4. B2B Labor Marketplaces 5. Infrastructure to Support B2B Marketplaces and Global Trade.

Discovery has moved online (Zillow, Trulia), but now transactions (Opendoor), co-living (Room, Asunder), lending, and yield-generation marketplaces are emerging. Transactions will become faster and more efficient.

Whether marketplace is demand- or supply-constrained; how tech-savvy users are; the effort required from supply or demand participants; how homogeneous target supply/demand is.

1. Shopping as Entertainment – browsing listings with no specific intent (e.g. Vinted, OLX); 2. Search – users know exactly what they want (e.g. Amazon, eBay); 3. Considered Purchases – when buyers need help choosing what to buy (e.g. high-value items, specialty goods)

The B2B marketplaces and the digitization of B2B supply chains is becoming a massive trend. Sub-1% penetrations in many verticals; lack of catalogs, online ordering, tracking, financing etc. SMB digitization is also a mega-trend. Supply chains moving out of China. Growth of labor marketplaces. Recommerce in B2B.

1. Verticalization of horizontals (Craigslist, eBay, Thumbtack, Upwork being replaced by verticals like Reverb, Block Renovation, Slice, Robin). 2. Reinvention of design: shifting from double-commit marketplaces (both sides choose) to ‘marketplace-pick’ (platform assigns supplier). 3. Rise of B2B marketplaces: bringing price transparency and efficiency to industries still dominated by rolodexes and opacity.

Recoup fully loaded CAC on a net contribution margin (CM2) basis in ~6 months; ~3x CAC on ~18 months; expect negative churn over time.

Poorly designed token incentives may attract users in it just for rewards (not quality), and tokenomics need to avoid undermining healthy cohorts. Like with discounts, you can degrade long-term behavior.

Onboarding too much supply early. Most marketplaces are demand-constrained; oversupply leads to churn and poor experience.

1. Team: must combine storytelling (to raise, hire, close deals) and analytical strength (unit economics). 2. Attractive business: strong TAM and good unit economics (CAC recoup in ~6 months, 3x CAC in 18 months, ideally negative churn). 3. Deal terms: reasonable given team, traction, and opportunity. 4. Thesis fit: does it align with FJ Labs’ thesis areas.

These are marketplaces for raw materials, heavy machinery, precision parts etc. They tend to have high Average Order Value (AOV), recurrence, and stickiness. Challenges include low take rates, difficulty switching incumbent behavior, and needing multiple revenue streams (SaaS, financing, logistics, insurance, etc.).

Vinted translated listings, shipping & payments integrated to become a pan-European marketplace. Ovoco (Lithuanian car parts) selling cross-border in Eastern Europe. Palm Street does live commerce for rare plants; WhatNot in collectibles.

Team (visionary, eloquent, can execute); Business (TAM and unit economics); Deal terms (fair for stage/traction/category); Thesis fit (future of work/food/real estate, making the world better).

Global. Roughly ~65% US/Canada, ~25% Western Europe/Nordics, ~10% Brazil/India, ~10% rest of world (e.g., Nigeria, Namibia, Chile, Vietnam, Philippines).

Use no-code/low-code or off-the-shelf tools when helpful; building an MVP today can be done cheaply (often <$50k, even <$25k).

Several:
• Cross-border commerce via auto-translation of listings & conversations plus better payments/logistics.
• Simplified listing process: AI can generate full listings (title, category, description, price) from minimal input.
• Improved listing quality (better images, background removal, optimized visuals) leading to higher sell-through.
• Productivity gains in customer support, marketing, dev workflows etc., through AI agents, auto-drafted replies, code generation tools etc.

Rise of off-premise ordering (Instacart, Uber Eats) complemented by verticals (Slice for pizza, Chowbus for Chinese, Stadium for catering). Growth of dark/cloud kitchens, virtual brands, and autonomous delivery.

“Friendshoring” refers to moving supply chains to geopolitically friendly countries. There is increasing tailwind due to geopolitics. India, Vietnam, Malaysia etc. are beneficiaries.

Recommerce refers to marketplaces for used goods or excess inventory being sold to other businesses (rather than only consumers). Example: Ghost enables brands to sell excess or outlet inventory to SMB stores in smaller batches.

Rebag, Herostuff, CollX are examples. Their AI can take just one photo to identify the item, suggest a price, generate the description, choose a category etc.

La Bourse aux Livres, a used book marketplace in France: you scan all your books, pack them, put on a shipping label they provide, ship to them; they immediately give you a 10% value coupon; they take care of selling. They make it work at a high take-rate due to the convenience.

Massive opportunity in digitizing B2B supply chains (often <1–5% digital). Five areas: (1) B2B marketplaces for inputs; (2) SMB enablement; (3) Friend-shoring; (4) B2B labor/talent marketplaces; (5) Infrastructure for B2B/global trade.

They get deals through founders they’ve backed before or through VCs who value their marketplace expertise. Their process involves a quick call to evaluate the team, business model, and market potential.

Traditional job boards (Indeed, ZipRecruiter) are weak arbitrage models. Vertical staffing marketplaces (RigUp, Trusted Health) create ongoing relationships, better economics, and stronger network effects.

Most B2B supply chains are still largely offline: pen & paper, no online catalog, no ordering, no tracking, no financing. There is sub-1% penetration of digital in many B2B supply chains; huge categories remain to be digitized.

Helping small businesses compete with large chains by handling what they don’t like doing. Examples: Odeko (coffee shops), Slice (pizzerias), Fresca (barber shops).

Helping small businesses with back-office tasks: managing inventory, POS solutions, websites, reviews, accounting etc. The marketplace often takes over tasks so that SMBs can focus on what they love doing. Monetization varies: margin, flat fees, small take rates.

Instead of RFQs and sorting through applications, the platform chooses the supplier for you (Uber choosing the driver, Comet picking the developer). This reduces friction and improves UX.

In contrast to older models where users chose suppliers, platforms now curate and assign suppliers (Uber picks the driver, Comet picks the developer, Miro picks the photographer). This reduces friction and boosts NPS.

In most industries B2B digital penetration is below 5%, and often below 1%. Many B2B categories still lack online catalogs, pricing, factory connectivity, ordering, payments, and order tracking.

Job boards (Indeed, ZipRecruiter) are arbitrage businesses with weak network effects. Staffing marketplaces (RigUp, Trusted Health) create ongoing relationships and network effects in verticals.

Start with highly curated, high-quality, engaged supply; bring enough demand to represent ~20% of their revenue (services) or ~20–25% sell-through (C2C). Then scale supply and demand in parallel (often hyper-local first).

Instead of generic horizontal platforms, verticals provide better UX and deeper integration (e.g., Reverb for music instruments, Slice for pizza shops, Robin for medical transcription).

Horizontal platforms (Craigslist, eBay) are being broken down into verticals like Reverb (music instruments), Thumbtack alternatives, Slice (pizzerias), Block Renovation (bathrooms). Vertical marketplaces create better, integrated experiences.

Digital freight forwarding, last-mile logistics, payment infrastructure, picking and packing warehouses, robotization/automation (e.g. warehouses, manufacturing), etc.

For used-goods C2C: ~20–25% probability of sale. For services/labor: represent ~20% of each supplier’s income before scaling supply further.

Execution, network effects, liquidity (getting buyers & sellers), matching efficiently, brand, delighting supply & demand and maintaining high NPS. Product itself is not a moat; many marketplace techs are replicable.

Around ~$150k/month GMV with ~15% take rate; gross margin ~60–70%; raising ~$3M at ~9–12M pre; aim for Series A at ~750k–$1M/month GMV with good unit economics.

Recoup fully loaded CAC on a net contribution margin (CM2) basis in ~6 months; 3x CAC on an ~18-month basis (or clear path with scale if early).

Innovation across models: iBuyers like Opendoor (instant cash offers, faster closings), Roofstock (yield-generating real estate), co-living startups, and short-stay brands building on top of Airbnb.

Effects are logarithmic: big gains improving wait times from 10→5→3 minutes; diminishing returns beyond ~3 minutes; explains why the space is not pure winner-takes-all.

Crypto-focused marketplaces are for trading crypto assets (like NFTs). Crypto-enabled marketplaces are non-crypto (or offline) platforms that use crypto mechanics; users don’t need to buy tokens to use them—they’re abstracted to maximize adoption.

Horizontal lenders (LendingClub, Funding Circle) are being verticalized. Example: Clearbanc links into Shopify and ad accounts to lend based on ROI—something large banks cannot do.

Start with the supply side first—specifically very high quality, engaged sellers. Be highly curated. Then test marketing channels for demand. Scale supply & demand in parallel. Ensure matching and delighting both sides.

Typically, you start by acquiring supply first, because suppliers are financially motivated to join. However, most marketplaces are demand-constrained.

Depending on setting, some parts of a marketplace can be decentralized with token mechanics, while core operations (quality, trust, customer service etc.) may remain centralized. Hybrid models are common.

Go narrow to go broad later if buyers purchase that SKU standalone; simplifies operations and improves quality (example logic given with vertical focus like only handbags before expanding).

B2B marketplaces. Many industries are <1–5% digital; trillion-dollar verticals (petrochemicals, steel, gravel, labor, etc.) need catalogs, factory connectivity, payments, tracking, insurance, financing.

We are still at the very beginning: only ~15% of commerce is online (25% in China). Education, healthcare, food, real estate, and other GDP sectors remain mostly offline.

Charge the more inelastic side or the side receiving the value; model varies by category (example given: buyer-side fees can unlock more supply).

Most B2B transactions are still offline, opaque, and relationship-driven. Marketplaces (e.g., freight forwarding, fish markets, scrap metal) are emerging to digitize and add transparency.

Because if they do a good job, they lose customers (users find a relationship and leave). Fabrice prefers businesses where good service increases usage over time, like Uber or staffing models.

Marketplaces unlock massive value through liquidity and transparency. They benefit from extraordinary feedback loops and network effects that create ‘winner takes all’ dynamics.

B2B is still mostly offline, with relationships and opacity dominating transactions. Younger, digital-native decision makers are pushing for Amazon-like user experiences in fragmented industries.

Early fundraising often depends on the founder’s ability to tell a compelling and convincing story.

Because B2B accounts for a large percentage of GDP, many categories have trillions of dollars in annual GMV potential, and the low penetration means long runway ahead for startups across verticals & geographies.

They bring liquidity and transparency to opaque, fragmented markets, unlocking massive value. They also benefit from feedback loops and network effects that drive winner-takes-all dynamics.

Yes. Marketplaces already have advantages: large selection, fulfillment infrastructure, customer support etc. Even if AI helps with discovery, those marketplaces with strong vertical dominance are likely to keep capturing most of the transaction value.