b2b-marketplace-launch-coach

Coach a founder through launching a B2B marketplace. B2B marketplaces are among the most attractive but most failed business models in venture: they have winner-take-most dynamics, large TAMs in fragmented industries, and durable defensibility — but ~80% of attempted B2B marketplaces fail to reach Series B. The reasons are knowable.

This skill is for founders building marketplaces where both sides are businesses (or where one side is a business and the other is a service provider/freelancer with business intent). For consumer marketplaces (Airbnb, DoorDash, Etsy), this is the wrong skill — different liquidity dynamics.

When to engage

Trigger when:

• "We're building a marketplace for [vertical] — buyers and sellers"
• "Faire for [X]" / "Flexport for [Y]" / "ServiceTitan for [Z]"
• "Should we charge commission or SaaS subscription?"
• "Chicken and egg — should we build supply or demand first?"
• "We have 20 buyers and 5 suppliers — is this enough liquidity?"
• "Our buyers are bypassing us after the first transaction"
• "Is this a marketplace or a directory or a SaaS tool?"

Don't engage when:

• It's a consumer marketplace (route to a different coach)
• It's a one-sided platform (B2B SaaS, not a marketplace)
• It's an "online catalog with payments" — that's distribution, not a marketplace

The first conversation: is this even a marketplace?

Many founders mislabel their idea. A marketplace has:

• Two distinct user types with different jobs-to-be-done
• Discovery as a core value prop (buyer finds new sellers, not just transacts with existing relationships)
• Liquidity as the moat (more participants make the platform more valuable)
• Repeated transactions between different counterparties

If the answer is "buyers come to find vendors they already know" — that's a directory or a CRM, not a marketplace. If the answer is "we charge per transaction but each customer only transacts with one vendor" — that's a payments/contracting tool, not a marketplace. If the answer is "we're a SaaS tool with a partner ecosystem" — that's a SaaS platform with referrals, not a marketplace.

Be willing to tell the founder they don't have a marketplace. Many "B2B marketplace" pitches are actually vertical SaaS or procurement tools with marketplace ambitions. The model is fine; the framing is misleading and will mislead the build.

Diagnostic intake

1. Who are the two sides? — Be specific. "SMB restaurants" + "food distributors" is different from "any business" + "any service provider".
2. What's the transaction? — Sourcing? Hiring? Procurement? Logistics? Consulting? Each has different dynamics.
3. Frequency? — One-time (M&A advisors), occasional (commercial real estate), recurring (food orders), continuous (fleet logistics)?
4. Average transaction value? — $100, $1K, $10K, $100K, $1M+? AOV drives everything: take rate, sales motion, defensibility.
5. Existing alternatives? — Direct relationships, RFPs, referral networks, Google search, existing marketplaces, brokers? You're stealing from these, not creating new behavior.
6. Stage? — Pre-launch, MVP with \x3C10 transactions, growth (100+ transactions/mo), scaling (1000+ transactions/mo)?
7. Vertical or horizontal? — Vertical (one industry deep) almost always beats horizontal in early stages.
8. Regulated? — Healthcare, legal, financial, government, alcohol, cannabis, defense have specific compliance moats (and traps).
9. Geographic scope? — Single city, single country, global? B2B marketplaces are usually national or global within a vertical.
10. Founder background? — Domain expertise in the vertical is 5x more important than marketplace experience for B2B specifically.

The B2B marketplace archetype taxonomy

Sort the company into one of these archetypes; each has different playbooks:

A. Vertical SaaS-enabled marketplace

Model: Start as SaaS for one side (usually supply); add demand-side as natural extension; transactions become the moat. Examples: Toast (POS for restaurants → supply chain orders), ServiceTitan (FSM for trades → marketplace ambitions), Faire (wholesale OS + marketplace). Strength: Strong supply-side lock-in via software; demand-side is naturally pulled in. Weakness: Slow to start (build SaaS first), expensive to build. Cold-start: Sell SaaS for 2-4 years, then layer marketplace on top.

B. Pure horizontal services marketplace

Model: Connect any business buyer with any service seller; rely on reviews/reputation. Examples: Upwork (freelancers), Toptal (vetted freelancers), Catalant (consultants). Strength: Large TAM if it works. Weakness: Brutal commoditization; race-to-bottom on price; high disintermediation. Cold-start: Concierge / vetted launch; build reputation systems early; geography or vertical entry point.

C. Vertical procurement marketplace

Model: Specific industry's procurement workflow (food, packaging, MRO, industrial supply). Examples: Faire (wholesale retail), Knowde (chemicals), Material Bank (architectural samples), Choco (restaurant procurement). Strength: Deep workflow integration; high switching cost once embedded. Weakness: Slow vertical-by-vertical expansion. Cold-start: Single vertical + single transaction type; manual matchmaking initially.

D. Vertical service marketplace

Model: Connect businesses with vertical-specific service providers (lawyers, accountants, marketing agencies, dev shops). Examples: Mayple (marketing agencies), Toptal (devs), LawTrades (legal), Pilot (bookkeeping). Strength: Vetted quality; clear SLA/standardization. Weakness: Service quality is hard to scale; disintermediation after first project is rampant. Cold-start: Concierge from day 1; aggressive vetting of supply; build reviews / case studies.

E. Logistics / supply-chain marketplace

Model: Connect shippers, carriers, brokers in transportation. Examples: Convoy (failed), Flexport (struggling), Uber Freight, Project44. Strength: Massive TAM; clear pain. Weakness: Capital-intensive, long sales cycles, race-to-bottom commoditization, brutal during downturns. Cold-start: One lane, one customer type, prove unit economics before scaling.

F. Aggregator marketplace

Model: Aggregate fragmented supply (small businesses) and present a unified buying experience. Examples: Joist (contractors → unified scheduling), GoSite (SMB tools). Strength: Buyers love the simplicity. Weakness: Supply-side commitment is hard; needs SaaS + marketplace combo. Cold-start: Sign up supply with SaaS first; transactions follow.

The cold-start playbook for B2B marketplaces

The hardest problem. B2C marketplaces solve cold-start with PR + viral loops + paid acquisition; B2B marketplaces don't have those tools.

Anti-pattern: launch with both sides simultaneously

Almost always fails. You spread effort thin, neither side gets enough density to retain, and you burn cash without either side experiencing the value.

The 5 viable B2B cold-start strategies

1. Single-player mode (build a tool one side uses without the other) Build software that one side uses regardless of the other side existing. Toast built POS that restaurants used standalone for years before becoming a marketplace. Faire built a wholesale-buyer dashboard before having sellers. When right: When you can build standalone value for one side. Software-heavy archetypes A and F. Cost: High ($500K-5M to build the software). Slow ($1-3 years to PMF). Why it works: The single-side product gives you a wedge into the side; transactions are downstream.

2. Concierge MVP (manual matchmaking)

Founder personally matches buyers and sellers — no software at first. Just a spreadsheet, phone calls, emails. When right: Services marketplaces, complex transactions, regulated verticals. Cost: Low ($0-100K). Slow scaling but high learning velocity. Why it works: You learn what buyers actually want before building the wrong product. Toptal did this for years.

3. City-by-city / vertical-by-vertical

Win one geographic or vertical slice completely; expand only after dominance. When right: Local services, vertical specialty marketplaces. Cost: Moderate. Slow but compounding. Why it works: Liquidity is a local phenomenon. NYC restaurant marketplace beats nationwide-thin marketplace.

4. Constraint-side first

Identify which side is structurally harder to acquire and start there. In most B2B marketplaces, demand (buyers) is the constraint — businesses are slow to change procurement habits. In some, supply is the constraint (specialized providers, regulated talent). When right: When you have a clear answer to "which side is harder". How to know: Run small experiments. Survey 50 of each side. Build a pre-launch waitlist.

5. SMB-first (avoid enterprise procurement complexity)

Start with SMBs (1-50 employees) who can self-serve. Move upmarket later. When right: Most B2B marketplaces. SMB sales cycles are 1-4 weeks; enterprise are 6-18 months. Why it works: You can iterate fast at SMB scale; SMB buyers are forgiving of MVP roughness. Risk: TAM may be too small; SMBs churn.

The rule of "100 transactions in one slice"

Before you scale geography or vertical, get to 100+ transactions per month in one slice (one city, one vertical, one customer-segment). Below 100, you don't have a marketplace — you have a manual matchmaking service. Don't expand geographically or by vertical before this milestone.

Pricing models — the economics

Commission only (% of transaction value)

Examples: Most consumer marketplaces, eBay, Faire (15-25% commission). Pros: Aligned with customer value; no friction to join. Cons: Take rate visible to buyers and sellers — disintermediation pressure; commission % gets squeezed over time. Take rate ranges: 5-10% (low-touch B2B procurement); 10-25% (services with vetting); 25-40% (managed services with quality moat); 40%+ (rare, requires unique value).

SaaS subscription + commission (hybrid)

Examples: Toast (POS SaaS + ordering commission), ServiceTitan (FSM SaaS + payment processing). Pros: Recurring revenue; SaaS reduces commission pressure; deeper integration. Cons: Two sales conversations; SaaS pricing must be justified independent of marketplace value. Best for: Archetypes A and F.

Lead-gen / per-introduction

Examples: HomeAdvisor, ZipRecruiter (B2B side), Bark. Pros: Easy to start; sellers pay only for leads. Cons: Race-to-bottom on lead quality; sellers eventually try to bypass. Take rate: $10-500 per qualified lead; effective take rate 1-5%.

Membership / subscription (both sides pay membership)

Examples: Catalant (paid by demand side), Mainspring (some categories). Pros: Selects for serious users; predictable revenue. Cons: High friction; small TAM; only works for niches with high willingness-to-pay.

Procurement / RFP fees

Examples: Coupa, Tradeshift (procurement-software-as-marketplace). Pros: High AOV per transaction; deep enterprise lock-in. Cons: Long sales cycles; complex implementation.

What take rate is achievable?

The answer depends on:

• Value-add per transaction: Pure intro = 1-5%. Vetting + payments + escrow + dispute = 15-30%. Full managed (you're the prime contractor) = 30-50%.
• Disintermediation risk: Higher = lower sustainable take rate. Recurring relationships disintermediate; one-time don't.
• Buyer's alternative: If they could find the same supplier on Google for free, 5%. If you're providing scarce supply, 25%+.
• Vertical norms: Real estate brokers take 3-6%, recruiting takes 20-25%, freelance platforms take 10-20%.

Rule of thumb: If founder says "we'll charge 30%+", probe hard on what they're delivering for that. 30%+ requires real value-add or scarce supply or both.

Unit economics that actually work

CAC by side

• Demand-side CAC: $50-500 for SMB; $500-5K for mid-market; $5K-50K for enterprise.
• Supply-side CAC: Often 2-5x demand CAC because supply has more options (existing customers, other marketplaces).
• Demand CAC payback: Should be \x3C12 months for SMB, \x3C18 months for mid-market, \x3C24 months for enterprise.

Take-rate × frequency × AOV = revenue per active relationship

Example: 15% take × 12 transactions/year × $5K AOV = $9K/year revenue per active buyer-seller relationship. For unit economics to work, CAC should be \x3C 50% of LTV (LTV = revenue × gross margin × retention years × #relationships).

Gross margin

• Pure software-only marketplace (no payments, no logistics): 80-90%.
• Marketplace with payments: 70-85% (after payment processing).
• Marketplace with logistics/fulfillment: 30-60% (depends on how much you take on).
• Managed services marketplace: 20-40% (you're paying for the service delivery).

The "good enough" benchmarks for B2B marketplace at Series A

• Net revenue retention >100%
• Take rate 10%+
• Frequency >3 transactions/year per active buyer
• Cohort retention M12 >70%
• LTV/CAC >3
• CAC payback \x3C18 months

The disintermediation problem

The defining failure mode of B2B marketplaces. Buyer and seller transact through you once; on the next transaction, they go direct.

Why it happens in B2B specifically

• Repeat transactions are common (food orders, cleaning service, contractor)
• Both sides are professionals with motivation to optimize
• The take rate is a visible cost
• Trust is built after one transaction; the marketplace's matchmaking value declines

Defenses

1. Payment as moat You hold escrow, manage invoicing, run net-terms financing. Disintermediation means losing the financial workflow. Faire offers Net 60 financing → buyers can't disintermediate without losing financing.

2. Software value beyond transaction You're the buyer's procurement system, the seller's CRM, the inventory tracker. Disintermediation means losing the software. ServiceTitan ties payment to FSM software.

3. Insurance / compliance / SLA You guarantee delivery, quality, compliance. Disintermediation means losing the guarantee. Toptal guarantees freelance quality; Pilot guarantees bookkeeping accuracy.

4. Discovery moat Buyer keeps coming back because they keep finding new sellers. Works when buyer needs are heterogeneous. Faire works because retailers want to discover new brands continuously.

5. Data moat You provide market intelligence, benchmarking, demand forecasting. Disintermediation means losing visibility. Convoy did this in freight (didn't save them).

6. Contract terms Hardcore: contractually require both sides to transact through you. Workable in some verticals (regulated supply); legally fragile in most. Don't lead with this.

The numbers

Healthy B2B marketplace: 60-80% of transactions on platform after relationship is established. Failing B2B marketplace: \x3C40% of transactions on platform; buyers and sellers know each other and transact directly. Survey both sides at 90 days post-first-transaction: "Have you transacted with [counterparty] outside the platform?" If yes >40%, you have a disintermediation crisis.

Regulated verticals: compliance as moat

Healthcare, legal, financial, government, alcohol, cannabis, defense — these have specific compliance requirements that are barriers but also moats.

What to build first

• Compliance documentation: HIPAA, SOC 2, PCI, state-by-state licensure tracking, KYC, AML.
• Trust signals: Certifications visible on listings, audit trails, dispute resolution.
• Vetting: Aggressive supply-side vetting in regulated verticals — you're effectively underwriting.

Where to spend

• Compliance team: 1-2 FTEs from day 1 in regulated verticals.
• Legal counsel: $50-150K/year retainer; never skimp on contracts and regulatory.
• Insurance: E&O, GL, cyber. $25-100K/year.
• Audit / certification: SOC 2 Type II ($25-75K). Vertical-specific (HITRUST for healthcare, FedRAMP for government).

The trap

Founders in regulated verticals over-invest in compliance before product-market fit. Compliance is necessary but not sufficient. Don't spend $300K on SOC 2 + HIPAA + PCI before you have 10 paying customers. Get conditional approvals, sign indemnifications, build the moat post-PMF.

"Managed marketplace" — the upgrade path

Most successful B2B marketplaces start open and become managed.

Open → vetted → managed → concierge progression

Open (anyone signs up, anyone transacts): cheap to scale, race-to-bottom on quality. Vetted (supply-side approval process): more trust, slower scaling. Managed (you facilitate transaction with SLA): higher take rate (20-30%), real moat. Concierge (you're effectively the prime contractor): highest take rate (30-50%), capital-intensive.

When to upgrade

• Open → Vetted: When supply quality is the #1 buyer complaint. Cost: 20-50% reduction in supply count, replaced by higher quality.
• Vetted → Managed: When transactions are still fragile (delivery, quality, payment) and you can take responsibility. Cost: build operations team, dispute resolution, SLA monitoring.
• Managed → Concierge: When you can guarantee outcomes for premium customers willing to pay 2-3x. Cost: dedicated operations, capital for working capital float.

The dual-tier strategy

Many mature B2B marketplaces run self-serve (open/vetted) + concierge (managed/concierge) tiers to capture both budget and premium buyers.

Network effects in B2B — much weaker than B2C

A mistake founders inherit from B2C is assuming "network effects = winner-takes-all = valuation premium". In B2B:

• Cross-side network effects (more buyers attract sellers and vice versa) are real but weaker than in B2C.
• Same-side network effects (more buyers attract more buyers) are usually negligible. Buyers don't compete with each other for marketplace attention.
• Data network effects (more transactions improve matching) are real but require structural design.
• Supply density network effects (when supply density past a threshold makes the marketplace meaningfully better) are real and powerful in local services.

The structural network effects in B2B marketplaces

• Trust / reputation systems: More transactions → better reviews → more transactions
• Pricing intelligence: More transactions → better benchmarking → buyers come back for pricing data
• Standardization: Many participants → standard contracts/terms → easier transactions
• Compliance reuse: Once you certify a supplier, all buyers benefit

What to invest in

• Reviews/reputation that buyers actually use (not vanity)
• Pricing transparency / benchmark data
• Standard contracts / payment terms
• One-time supplier onboarding that all buyers benefit from

Funding dynamics

Seed

What investors look for: Founder/market fit, vertical depth, conviction on which side is the constraint. Some traction (10-100 transactions, 5-20 active relationships, $100K-$1M revenue run-rate possible). Round size: $1.5-5M. Valuation $10-25M.

Series A

What investors look for: $1-5M ARR, 100+ active monthly buyers, evidence of cohort retention >70% at M6, take rate trajectory positive. Round size: $8-20M. Valuation $40-120M. Common stall: Most B2B marketplaces stall at $1-3M ARR with weak cohort retention.

Series B

What investors look for: $10-25M ARR, NRR >100%, gross margin clarity, take-rate sustainability, and (often) evidence of category-leadership in the vertical. Round size: $25-60M. Valuation $200-500M. Common stall: Cohort retention turns negative, or unit economics don't scale (CAC rises faster than LTV expands).

After Series B

Funding is hard. B2B marketplaces have a "valley of death" between Series B and growth-equity-stage profitability. Many great B2B marketplaces (Convoy, others) failed here. Key question: Are you path-to-profitability before next round? If not, you need extraordinary growth.

Common founder mistakes

1. Treating it as a B2C marketplace — "we'll just do paid acquisition like Airbnb did" — doesn't work in B2B.
2. Building both sides equally — usually fails.
3. Pricing without value-add — taking 20%+ when you only do introductions.
4. Ignoring disintermediation until it's a crisis — losing 60%+ of transactions to off-platform by year 2.
5. Over-investing in compliance pre-PMF — burns 12-18 months and $500K-2M on certifications before knowing the product.
6. Vertical-creep too early — expanding from food to fashion before owning food.
7. Geographic-creep too early — going national before owning one city.
8. Not measuring frequency / cohort retention — vanity GMV growth masks bad unit economics.
9. Building the marketplace tech before having transactions — you don't know what to build.
10. Hiring sales before having a repeatable selling motion — sales scales pain, doesn't fix it.

Output format

Always produce:

• Marketplace archetype identification: which of A-F
• Cold-start strategy recommendation: one of the 5
• Pricing model recommendation: with specific take-rate range
• Unit-economics target: CAC, LTV, payback, take rate, frequency
• Disintermediation defense plan: 2-3 mechanisms appropriate to the model
• Network effect plan: which structural mechanisms to invest in
• First-100-transactions plan: explicit milestones
• Failure-mode flags: which of the 10 mistakes is highest-risk
• Funding stage map: what to prove for next round

Anti-patterns

• Don't recommend going horizontal early — vertical first, almost always
• Don't recommend SaaS + marketplace simultaneously without 18 months runway
• Don't recommend skipping concierge phase — almost every successful B2B marketplace did concierge for the first 6-18 months
• Don't accept "GMV" as a primary metric pre-PMF — focus on frequency, cohort retention, take rate
• Don't recommend marketplace launch in highly-fragmented vertical without supply-side strategy

What "great" looks like at 12 / 24 / 36 months

Year 1: 100+ transactions/month in one slice; $200K-1M ARR; cohort retention M6 >75%; take rate 10%+; one geographic/vertical "owned". Year 2: 1000+ transactions/month; $2-5M ARR; expansion into 2-3 slices; disintermediation rate \x3C30%; clear unit-economics path. Year 3: 10,000+ transactions/month; $10-20M ARR; recognized category leader; gross margin 60%+; raise Series B.

A bad year-1 looks like:

• Plenty of "registered users" but low transaction frequency
• Take rate below 5% with no path to expand it
• Cohort retention M6 \x3C50%
• Spreading thin across 3+ verticals/cities
• High GMV growth via low-margin, low-defensibility transactions

Coach toward the first picture, away from the second.