A UK founder we spoke to last month wanted to build "Amazon for kitchenware." His pitch deck was strong. His software budget was £180k. His seller acquisition budget was £0. His plan for onboarding the first 50 sellers was "we will build the platform and they will come." Six months later, he had a beautiful marketplace with 3 sellers and no transactions.
That is the marketplace conversation across UK and US founders in 2026. Marketplace software is more capable than ever (Sharetribe, white-label platforms, custom builds all viable). The hard part has never been software. Seller acquisition, trust and safety operations, and the cold-start chicken-and-egg problem are what kill marketplaces. Founders that plan for the operational reality survive; founders that only plan the software do not.
This article is a candid guide for founders scoping multi-vendor marketplace development. What marketplace actually involves beyond software. Build options and their trade-offs. Real cost bands. The cold-start problem and how to survive it. Payment layer choices. Written for founders considering the build, not those already deep in operations.
What a Multi-Vendor Marketplace Actually Involves
Twelve functions beyond software.
Seller acquisition and onboarding. How you find, recruit, verify, and enable sellers. Manual initially; can be systematised later. Almost always the biggest operational cost.
KYC/AML compliance. Know Your Customer and Anti-Money Laundering checks per seller. Legally required in most jurisdictions. Handled by payment provider (Stripe Connect, Adyen) or specialist (Onfido, Sumsub).
Payment splitting and payouts. Split payment between marketplace, seller, and fees per transaction. Automated payouts on schedule. Handled by Stripe Connect, Adyen MarketPay, or Mangopay.
Commission and fee structure. Percentage, flat, or hybrid fee model. Payment processing markup. Subscription vs transaction fees. Structure affects seller economics and marketplace revenue.
Ratings and reputation system. Buyer rates seller after transaction. Seller reputation influences buyer trust and search ranking. Trust primitive that separates marketplaces from bulletin boards.
Dispute resolution process. When transactions go wrong, marketplace mediates. Formal process, escalation path, refund policy, seller strike system. Operational function requiring people.
Trust and safety operations. Detecting fake listings, fraudulent sellers, abusive buyers, prohibited content. Growing as the platform scales.
Category management. Product categorisation, taxonomy, attributes. Enables search and browsing. Requires domain expertise per category.
Search relevance. How buyers find sellers/products. Elasticsearch, Algolia, or Meilisearch typical. Requires ongoing tuning.
Fraud detection. Card fraud, seller fraud, promo abuse. Growing operational function.
Legal terms and marketplace liability. Marketplace terms, seller agreement, buyer terms, liability allocation. Requires marketplace-specialist legal counsel.
Tax and VAT handling per seller. Marketplace facilitator obligations vary by jurisdiction. UK VAT, EU digital services, US sales tax. Complexity grows with geographic coverage.
Sharetribe vs White-Label vs Custom Build
Three technical patterns.
Sharetribe. No-code/low-code marketplace platform. £99-£299/month base plus setup. Fast MVP (2-4 weeks). Best for founders validating concept before committing capital. Limited customisation; you fit their model.
White-label (Yo!Kart, Arcadier, Marketplacer, others). Licensed platform with more customisation than Sharetribe. £15k-£100k license plus setup and ongoing. Best for founders wanting more control than Sharetribe but not full custom.
Custom build. Full control, full cost. £120k-£500k for MVP, £500k-£2m for scale-ready. Best for marketplaces with unusual workflow, unique commission models, or long-term differentiation requirements.
Which to choose.
Concept validation and small budget: Sharetribe
Small-to-medium marketplace with standard workflow: white-label
Large or differentiated marketplace with growth ambition: custom
Most successful marketplaces start with Sharetribe or white-label for validation, then migrate to custom once concept is proven and unit economics work. Migrating too early wastes capital; migrating too late loses to competitors.
The Cold Start Problem (What Kills Most Marketplaces)
Marketplaces die at cold start, not at software.
The problem. No sellers means no buyers. No buyers means no sellers join. Marketplace launches empty and stays empty. This is why 90 percent of marketplace startups fail regardless of software quality.
Solution 1: Single-sided seed. Recruit sellers first with subsidy, exclusivity, or manual work. Marketplace has inventory when buyers arrive. Requires founder to do sales work, not just build software. Airbnb photographed apartments manually; Etsy did craft fair recruitment.
Solution 2: Niche vertical. Small enough vertical that 20-50 sellers cover it. Founder can onboard manually. Growth from niche outward. Amazon started with books before adding categories.
Solution 3: Platform hijacking. Integrate with existing platform's sellers. Booking.com fed off existing hotel booking systems. TaskRabbit fed off Craigslist initially. Reduces cold start.
What does not work. "Build it and they will come." "We will do a big launch campaign." "We will find sellers after software is done." These have killed thousands of marketplaces.
The McKinsey Digital insights marketplace research consistently shows cold start solving as the single strongest predictor of marketplace survival past year one.
Real 2026 Cost Bands
MVP options.
Sharetribe MVP: £5k-£25k over 2-4 months (mostly setup and customisation)
Custom MVP: £80k-£200k over 4-8 months
White-label MVP: £30k-£150k over 3-6 months
Scale-ready builds.
Custom scale-ready marketplace: £250k-£800k over 8-14 months
White-label enterprise: £100k-£400k over 6-12 months
Enterprise marketplace.
Custom enterprise: £800k-£2m plus over 12-24 months
Includes advanced features (multi-currency, multi-language, complex commission models, enterprise search)
Ongoing operational cost (independent of software).
Small marketplace: £5k-£20k monthly (small support team)
Medium marketplace: £30k-£100k monthly (support, trust and safety, category management)
Large marketplace: £150k plus monthly (multiple operational functions at scale)
Software cost is the smaller number in each case. Operational cost dominates over 12+ months. Founders that budget only for software underspend on the actual expensive parts.
Stripe Connect and the Payment Layer
Payment layer is the single most complex piece of marketplace software.
Stripe Connect. Global standard for marketplace payment splitting. Handles payment processing, KYC, payouts, tax reporting. Available in most major markets. Charges £0.25 per payout plus 0.25 percent additional on transaction on top of standard Stripe processing fees. See Stripe Connect for detailed capability.
Alternatives.
Adyen MarketPay: strong for enterprise marketplaces, better rates at scale
Mangopay: EU-focused, e-money license, good for European marketplaces
Custom payment splitting: rare, expensive, high compliance burden
What Stripe Connect handles.
Card acceptance
Payment splitting (marketplace fee + seller amount)
Seller onboarding (KYC per seller)
Payouts on schedule
Tax form generation
Refund handling with seller pull-back
What it does not handle.
Dispute mediation (marketplace responsibility)
Refund policy enforcement (marketplace responsibility)
Advanced fraud detection beyond Stripe Radar (marketplace or third-party)
For most 2026 marketplaces, Stripe Connect is the default. Alternatives make sense for specific regulatory or geographic needs.
What We Learned Building for E-commerce
We built TrackVid, a video proof and claim management platform for Indian ecommerce sellers including some who operate their own marketplaces. Two lessons transfer to marketplace founders.
Operational function scaling drives most cost after year one. Trust and safety, dispute resolution, seller support, and category management scale with transaction volume, not with software complexity. Founders that under-budget these grow into operational chaos. Budget realistically for the human functions from day one.
Trust primitives (ratings, dispute resolution, video proof) are the compounding differentiators. Software features are competed away. Trust primitives compound over years. TrackVid gets used by marketplace sellers precisely because platform trust primitives are insufficient; video proof supplements what the marketplace should provide but does not.
You can see our shipped work at our portfolio. If you want a candid conversation about your marketplace build, book a marketplace scoping call with WhiteStone.
Common Failure Modes
Three failure modes we see repeatedly.
Building software before solving cold start. Beautiful marketplace launches with 3 sellers and no transactions. Software was the easy part; seller acquisition was the actual challenge.
Skipping operational function budget. Founder budgets £180k for software, £0 for operations. Software delivers; operations do not exist; marketplace fails at first serious dispute.
Starting with custom build for validation. Founder commits £250k to custom build before validating concept. Sharetribe validation would have cost £15k. Six months of custom build wasted when concept turns out not to work.
Frequently Asked Questions
Sharetribe or custom multi-vendor marketplace?
Sharetribe for concept validation and small budget (£99-£299/month plus £5k-£25k setup, 2-4 months to MVP). Custom for marketplaces with unusual workflow, unique commission models, or long-term growth ambition (£120k-£500k MVP, 4-8 months). Most founders benefit from Sharetribe validation before committing to custom.
How is commission calculated at scale?
Standard approach: percentage of gross merchandise value (5-20 percent typical), sometimes plus flat listing fee or subscription. Advanced marketplaces use tiered commission (lower rate for high-volume sellers), category-specific rates, or performance-based (lower rate for sellers with high ratings). Structure affects seller economics and marketplace revenue materially.
How do you onboard sellers safely?
KYC via Stripe Connect, Adyen, or specialist providers (Onfido, Sumsub). Manual review of first listings. Trust badge or verification tier system. Rating system for buyer feedback. Strike system for policy violations. Trust and safety team scales with volume.
What is the cold start problem for marketplaces?
No sellers means no buyers means no sellers join. Kills 90 percent of marketplace startups regardless of software. Solutions: single-sided seed (recruit sellers first with subsidy or manual work), niche vertical (small enough to onboard 20-50 sellers manually), or platform hijacking (integrate with existing platform's sellers). Software does not solve cold start; sales work does.
How much does a multi-vendor marketplace cost to build?
Sharetribe MVP £5k-£25k. Custom MVP £80k-£200k. Scale-ready custom £250k-£800k. Enterprise marketplace £800k-£2m plus. Add £5k-£150k monthly ongoing operational cost depending on scale. Software cost is the smaller number over 12+ months; operational cost dominates.
The One Thing to Remember
Marketplaces die at cold start, not at software. The founders that survive year one are the ones who solved seller acquisition before writing production code. Sharetribe or white-label validation followed by custom scale-up beats custom-from-day-one for most 2026 marketplace bets. Payment layer (Stripe Connect for most cases) matters more than founders realise. Operational function budget (trust and safety, dispute resolution, seller support) is where the money actually goes over 12+ months.
If you want a candid conversation about your specific marketplace build, browse our custom software development services or come to the call.
