--- # === IDENTITY === id: business/strategy/business-model-selection/2026 canonical_question: "How to select business model — subscription vs transactional vs marketplace vs platform vs freemium?" aliases: - "revenue model selection framework" - "subscription vs transactional vs marketplace comparison" - "which business model should I use" - "business model decision tree" - "freemium vs subscription vs marketplace decision" entity_type: decision_framework domain: business > strategy > Business Model Selection region: global jurisdiction: global temporal_scope: 2024-2026 # === VERIFICATION === last_verified: 2026-03-10 confidence: 0.85 version: 1.0 first_published: 2026-03-10 # === TEMPORAL VALIDITY === temporal_validity: status: evolving last_breaking_change: "Usage-based and hybrid models gaining rapid adoption across SaaS and marketplace sectors (2024-2025)" next_review: 2026-09-06 change_sensitivity: medium # === CONSTRAINTS === constraints: - "Covers five primary digital business models only — subscription, transactional, marketplace, platform, and freemium. Does not address advertising, licensing, or hardware-bundled models." - "Assumes digital-first or software product. Physical goods businesses need additional supply chain and inventory considerations." - "Business model selection is a two-way door early on but becomes increasingly irreversible as unit economics, infrastructure, and customer expectations solidify (typically after 12-18 months)." - "Marketplace and platform models require solving the chicken-and-egg problem before revenue materializes — runway must account for 12-24 months of pre-revenue operations." - "Multiple stakeholders (product, finance, engineering, sales) must align on the model — misalignment creates execution friction that no model can overcome." # === SKIP CONDITIONS === skip_this_unit_if: - condition: "User already chose a model and needs pricing strategy help" use_instead: "business/pricing/saas-pricing-models-comparison/2026" - condition: "User needs SaaS-specific unit economics benchmarks" use_instead: "finance/industry-benchmarks/saas-industry-benchmarks-2026/2026" - condition: "User is evaluating marketplace take rates specifically" use_instead: "business/industry-benchmarks/marketplace-industry-benchmarks-2026/2026" # === AGENT HINTS === inputs_needed: - key: "product_type" question: "What type of product or service are you offering?" type: choice options: ["SaaS / software tool", "Marketplace connecting buyers and sellers", "Platform enabling third-party development", "Digital content or media", "Physical goods with digital layer"] - key: "usage_frequency" question: "How often will customers use or need the product?" type: choice options: ["Daily or continuous (high frequency)", "Weekly or periodic (moderate frequency)", "Occasional or one-time (low frequency)", "Varies by customer segment"] - key: "marginal_cost" question: "What is the marginal cost of serving one additional customer?" type: choice options: ["Near zero (software, digital content)", "Low but non-trivial (cloud compute, storage)", "Moderate (requires human support or fulfillment)", "High (physical goods, professional services)"] - key: "tam_size" question: "How large is the total addressable market?" type: choice options: ["Very large (>10M potential users)", "Large (1-10M potential users)", "Medium (100K-1M potential users)", "Small or niche (<100K potential users)"] - key: "network_effects" question: "Does the product become more valuable as more users join?" type: choice options: ["Strong direct network effects (each user adds value for all)", "Strong indirect network effects (more supply attracts more demand)", "Weak or no network effects", "Uncertain"] # === DISTRIBUTION === canonical_source: "https://knowledgelib.io/business/strategy/business-model-selection/2026" suggested_citation: "Source: knowledgelib.io — AI Knowledge Library (verified 2026-03-10)" # === RELATED UNITS === related_kos: depends_on: [] leads_to: - id: "business/pricing/saas-pricing-models-comparison/2026" label: "B2B SaaS pricing models compared — per-seat, usage-based, flat-rate and freemium" - id: "finance/industry-benchmarks/saas-industry-benchmarks-2026/2026" label: "SaaS industry benchmarks 2026 — CAC, LTV:CAC, NRR, churn, gross margin, Rule of 40 by segment" related_to: - id: "business/build-vs-buy/build-vs-buy-vs-partner-decision-tree/2026" label: "Build vs Buy vs Partner Decision Tree — technology capability sourcing" often_confused_with: [] alternative_to: [] # === SOURCES === sources: - id: src1 title: "Subscription Model vs. Transaction Model: A Comparison" author: SubscriptionFlow url: https://www.subscriptionflow.com/2023/06/subscription-model-vs-transaction-model/ type: technical_blog published: 2023-06-15 reliability: moderate_high - id: src2 title: "Marketplace business models — how to choose the right one" author: Sharetribe url: https://www.sharetribe.com/academy/how-to-choose-the-right-business-model-for-your-marketplace/ type: technical_blog published: 2024-08-01 reliability: high - id: src3 title: "Why Some Platforms Thrive and Others Don't" author: Harvard Business Review url: https://hbr.org/2019/01/why-some-platforms-thrive-and-others-dont type: industry_report published: 2019-01-29 reliability: authoritative - id: src4 title: "The Network Effects Manual: 16 Different Network Effects" author: NFX url: https://www.nfx.com/post/network-effects-manual type: industry_report published: 2024-03-01 reliability: high - id: src5 title: "The Freemium Model: When it Works (and when it doesn't)" author: Goji Labs url: https://gojilabs.com/blog/the-freemium-model-when-it-works-and-when-it-doesnt/ type: technical_blog published: 2024-11-01 reliability: moderate_high - id: src6 title: "Subscription vs Transaction Model | GoCardless" author: GoCardless url: https://gocardless.com/guides/posts/subscription-model-vs-transaction-model/ type: technical_blog published: 2024-05-01 reliability: moderate_high - id: src7 title: "How subscription business models work" author: Stripe url: https://stripe.com/resources/more/subscription-business-models-101-types-of-models-how-they-work-and-how-to-choose-one type: official_docs published: 2024-09-01 reliability: high --- # Business Model Selection Decision Framework ## Summary This framework helps founders and product leaders choose between five primary digital business models — subscription, transactional, marketplace, platform, and freemium — based on product characteristics, usage patterns, marginal costs, market size, and network effects. The default recommendation for most SaaS products is subscription with usage-based tiers, because it provides predictable recurring revenue and higher customer lifetime value than transactional models. [src1] However, marketplace and platform models can generate significantly higher enterprise value when network effects are strong, as network effects account for approximately 70% of the value created by technology companies since 1994. [src4] ## Constraints - This framework covers five digital business models only: subscription, transactional, marketplace, platform, and freemium. Advertising-supported, licensing, and hardware-bundled models are out of scope. - The choice is a two-way door early in a product's life but becomes increasingly irreversible after 12-18 months as infrastructure, pricing expectations, and customer contracts calcify. - Marketplace and platform models face the cold-start problem — they require critical mass on at least one side before delivering value, which typically demands 12-24 months of pre-revenue investment. [src2] - Freemium requires a very large total addressable market (TAM) because only 2-5% of free users convert to paid. [src5] - Hybrid models (e.g., subscription + marketplace, freemium + usage-based) are increasingly common and often outperform pure models, but add operational complexity. ## Decision Inputs | Input | Why It Matters | How to Assess | |-------|---------------|---------------| | Product type | Software tools favor subscription; connecting buyers/sellers favors marketplace | Ask what the product does and who it serves | | Usage frequency | High-frequency use supports subscription; low-frequency favors transactional | Ask how often a typical customer needs the product | | Marginal cost per customer | Near-zero marginal cost enables freemium; high marginal cost eliminates it | Ask about infrastructure, support, and fulfillment costs per additional user | | Total addressable market (TAM) | Freemium requires >1M potential users; niche markets favor subscription or transactional | Ask about target market size and segment | | Network effects potential | Strong network effects favor marketplace/platform; no network effects favor subscription/transactional | Ask whether the product becomes more valuable as more users join | | Revenue predictability need | High need for predictable revenue favors subscription; tolerance for variability enables transactional or marketplace | Ask about investor expectations, burn rate, and planning needs | ## Decision Tree ``` START — Which business model fits best? ├── Does the product connect independent buyers and sellers? │ ├── YES — Are transactions the core value (not just connections)? │ │ ├── YES → EVALUATE: Marketplace (commission-based) │ │ │ ├── Can you solve the chicken-and-egg problem? │ │ │ │ ├── YES + Have 18-24 month runway → RECOMMEND: Marketplace │ │ │ │ │ Reason: Transaction facilitation is the product │ │ │ │ │ Constraint: Need critical mass on both sides │ │ │ │ └── NO → RECOMMEND: Subscription SaaS (tool for one side) │ │ │ │ Reason: Build value for one side first, add marketplace later │ │ │ └── Is the average transaction value high enough for commission? │ │ │ ├── YES (>$50 AOV) → Commission model viable │ │ │ └── NO (<$50 AOV) → Consider subscription or lead-gen model │ │ └── NO (connections only) → RECOMMEND: Subscription or lead-gen │ │ Reason: Low transaction value makes commission unviable │ └── NO │ ├── Does the product become more valuable with more users? │ │ ├── STRONG network effects │ │ │ ├── Can third parties build on your product? │ │ │ │ ├── YES → RECOMMEND: Platform (rev share + API fees) │ │ │ │ │ Reason: Third-party ecosystem multiplies value │ │ │ │ │ Constraint: Requires API investment and developer relations │ │ │ │ └── NO → RECOMMEND: Subscription with network tiers │ │ │ │ Reason: Network effects drive retention in subscription │ │ │ └── ... │ │ ├── WEAK or NO network effects │ │ │ ├── Is usage frequent (daily/weekly)? │ │ │ │ ├── YES → RECOMMEND: Subscription │ │ │ │ │ Reason: Frequent use justifies recurring payment │ │ │ │ └── NO → RECOMMEND: Transactional (pay-per-use) │ │ │ │ Reason: Infrequent users resist subscription commitments │ │ │ └── ... │ │ └── UNCERTAIN → Default to Subscription, add usage tracking │ └── Is the marginal cost near zero AND TAM > 1M users? │ ├── YES on both → EVALUATE: Freemium │ │ ├── Can you create a clear free/paid boundary? │ │ │ ├── YES → RECOMMEND: Freemium + premium tiers │ │ │ │ Reason: Free tier drives adoption, premium captures value │ │ │ │ Constraint: Must sustain 95-98% non-paying users │ │ │ └── NO → RECOMMEND: Free trial + subscription │ │ │ Reason: No clear boundary = free tier cannibalizes paid │ │ └── ... │ └── NO → RECOMMEND: Subscription or transactional │ Reason: Freemium requires scale and low marginal cost ├── OVERRIDE CONDITIONS (check these regardless of tree path): │ ├── Investor expects predictable MRR/ARR → Subscription component required │ ├── Regulatory requirement for per-transaction pricing → Transactional required │ ├── Product is infrastructure/API → Usage-based pricing preferred │ └── TAM < 100K users → Avoid freemium (insufficient conversion volume) └── DEFAULT (if inputs are ambiguous): └── RECOMMEND: Subscription with usage-based tiers Reason: Most forgiving model — predictable revenue, adjustable pricing, works across market sizes, and can layer on marketplace or freemium later ``` ## Options Comparison | Factor | Subscription | Transactional | Marketplace | Platform | Freemium | |--------|-------------|---------------|-------------|----------|----------| | **Revenue predictability** | High (recurring MRR/ARR) | Low (variable per period) | Medium (scales with GMV) | Medium-High (recurring + variable) | Low until conversion matures | | **Time to first revenue** | 1-3 months | Immediate | 6-18 months | 12-24 months | 3-6 months (from paid tier) | | **Customer acquisition cost** | Moderate ($200-700 SaaS avg) | Low per transaction | High (must acquire both sides) | Very high (must attract developers) | Low (free tier is the funnel) | | **Customer lifetime value** | High (multi-year retention) | Low (per-transaction) | High (if liquidity achieved) | Very high (ecosystem lock-in) | Variable (2-5% convert) | | **Scalability** | Linear with sales capacity | Linear with demand | Exponential with network effects | Exponential with ecosystem | Exponential if TAM is large | | **Best when** | Product is used frequently, delivers ongoing value | Usage is infrequent or one-time, high per-use value | Connecting fragmented supply and demand | Third parties can extend your product | TAM is very large, marginal cost near zero | | **Worst when** | Usage is sporadic and unpredictable | Customer needs ongoing access | Cannot solve cold-start problem | No developer interest in building on top | Marginal cost is non-trivial, TAM is small | | **Hidden costs** | Churn management, feature bloat pressure | Demand generation, revenue volatility | Fraud, trust/safety, supply quality | Developer relations, API maintenance | Supporting 95%+ non-paying users | [src1, src2, src3, src5] ## Decision Logic ### If product connects buyers and sellers AND transaction facilitation is core value AND 18+ month runway exists → **Marketplace (commission-based)**. The 51% of top marketplaces use commission as their primary revenue model because it aligns platform incentives with seller success. The commission rate should be 10-30% depending on value delivered and competitive alternatives. [src2] ### If product is a SaaS tool with frequent usage AND no strong network effects → **Subscription**. Subscription models produce higher long-term customer lifetime value than transactional models because they create ongoing engagement, enable upselling, and provide richer customer data for product improvement. Average B2B SaaS churn benchmarks are 3.5-5% annually. [src1, src7] ### If product has strong network effects AND third parties can build on it → **Platform (API + rev share)**. Platform business models generate the highest enterprise value when network effects take hold, but they require the longest runway and the highest upfront investment. Five critical dynamics must be evaluated: network effects, clustering, disintermediation risk, multi-homing vulnerability, and bridging potential. [src3] ### If TAM > 1M users AND marginal cost per user is near zero AND clear free/paid feature boundary exists → **Freemium + premium tiers**. Freemium works when the free tier demonstrates enough value to drive adoption while leaving enough premium features to motivate 2-5% conversion. Conversion below 2% signals the free tier is too generous or the premium tier is not compelling enough. [src5] ### If usage is infrequent or one-time AND customers resist subscription commitments → **Transactional (pay-per-use)**. High per-transaction value and infrequent use make transactional models more attractive than subscription. Customers who only need a product 2-3 times per year will not pay monthly fees. [src6] ### If early stage with uncertain product-market fit → **Start with subscription, add complexity later**. Subscription is the most forgiving starting model because it provides predictable revenue, is easy to price and adjust, and creates a direct customer relationship. Marketplace, platform, and freemium layers can be added once PMF is validated. [src7] ### Default recommendation → **Subscription with usage-based tiers**. When inputs are ambiguous, subscription with usage-based pricing components is the lowest-risk choice. It provides revenue predictability while allowing pricing to scale with customer value. Most SaaS companies begin with flat subscription and add usage-based components at Series A or B stage. [src7] ## Anti-Patterns ### Wrong: Choosing freemium without sufficient market size Companies launch a freemium model targeting a niche market of 50-100K potential users. With 2-5% conversion rates, this yields only 1,000-5,000 paying customers — often insufficient to sustain operations. Meanwhile, the company incurs costs supporting tens of thousands of non-paying users. [src5] ### Correct: Reserve freemium for markets with 1M+ potential users Freemium requires massive top-of-funnel volume to generate meaningful paid conversion. For niche markets, use a 14-30 day free trial with a subscription model instead — this delivers the same try-before-you-buy experience without the ongoing cost of free users. [src5] ### Wrong: Launching a marketplace without solving the cold-start problem first Founders build a two-sided marketplace and launch to zero supply and zero demand, expecting both sides to arrive simultaneously. Without critical mass, the marketplace delivers no value, and users churn before network effects can form. [src2] ### Correct: Concentrate on one side first, then expand Successful marketplaces typically start by providing standalone value to one side (usually supply). Sharetribe research shows that marketplaces that achieve liquidity on one side first have significantly higher survival rates than those attempting simultaneous two-sided launch. [src2] ### Wrong: Defaulting to subscription for everything Teams choose subscription because it is the most familiar model, even when usage patterns do not support recurring payment. Customers who use a product 2-3 times per year resent monthly charges and churn quickly, destroying the revenue predictability that made subscription attractive in the first place. [src6] ### Correct: Match the model to actual usage frequency Audit expected usage patterns before selecting a model. High-frequency use (daily/weekly) supports subscription. Low-frequency use (monthly or less) supports transactional or pay-per-use. Variable use supports usage-based pricing with a small base subscription. [src1] ## Cost Benchmarks | Scenario | Subscription | Transactional | Marketplace | Platform | Freemium | |----------|-------------|---------------|-------------|----------|----------| | **CAC (B2B)** | $200-700 | $50-200/sale | $100-500/side | $500-2,000/developer | $10-50 (free), $200-500 (convert) | | **CAC (B2C)** | $50-150 | $5-50/sale | $20-100/side | $100-500/developer | $1-10 (free), $50-150 (convert) | | **Payback period** | 6-18 months | Immediate | 12-36 months | 18-36 months | 12-24 months | | **Gross margin** | 70-85% | 40-70% | 60-80% (net revenue) | 70-90% | 70-85% (on paid users) | | **Annual churn (B2B)** | 3.5-5% | N/A (per transaction) | 15-25% (supply side) | 10-20% (developers) | 5-8% (paid users) | | **LTV:CAC target** | 3:1+ | 2:1+ per cohort | 4:1+ (long payback) | 5:1+ (high upfront cost) | 3:1+ (on paid users only) | **Hidden cost multipliers**: Add 15-25% for payment processing and billing infrastructure. Marketplace models add 10-20% for trust/safety and dispute resolution. Platform models add 25-40% for developer relations and API maintenance. Freemium models must budget for serving non-paying users — typically 60-80% of infrastructure costs serve free-tier users. [src5, src7] ## When This Matters Fetch when a founder, product leader, or strategy consultant is choosing a business model for a new product, evaluating whether to change an existing business model, or comparing revenue model options. Especially relevant at pre-seed/seed stage, before product-market fit, or when pivoting from one model to another. ## Related Units - [SaaS Pricing Models Comparison](/finance/saas-benchmarks/saas-pricing-models-comparison/2026) - [SaaS Unit Economics Benchmarks](/finance/saas-benchmarks/saas-unit-economics-benchmarks/2026) - [Build vs Buy vs Partner Decision Tree](/business/build-vs-buy/build-vs-buy-vs-partner-decision-tree/2026)