Business Models for Software Startups
How you make money shapes everything about your company. Here are the business models that work for software startups.
Your business model is how you create and capture value. It determines how you price, sell, hire, and build. Choosing the right model for your market is one of the most important strategic decisions you’ll make.
The Major Software Business Models
SaaS (Software as a Service)
Customers pay recurring subscription fees to access software hosted in the cloud.
Revenue: Monthly or annual subscriptions
Examples: Salesforce, Slack, Notion, Figma
Characteristics:
•
Predictable recurring revenue
•
Customer success critical (churn kills)
•
Often land-and-expand model
•
Requires ongoing product development
Key metrics:
•
MRR/ARR (Monthly/Annual Recurring Revenue)
•
Churn rate
•
Net Revenue Retention
•
CAC/LTV ratio
Works when:
•
Software provides ongoing value
•
Customers need regular updates
•
Central hosting provides benefits
•
You can build switching costs
Transactional/Usage-Based
Customers pay per transaction or based on usage (API calls, messages sent, compute used).
Revenue: Variable based on customer activity
Examples: Stripe (% of transaction), Twilio (per message), AWS (per compute hour)
Characteristics:
•
Revenue scales with customer success
•
Low barrier to entry
•
Revenue can be volatile
•
Heavy users subsidize light users
Key metrics:
•
Gross Transaction Volume
•
Take rate (your cut)
•
Usage growth
•
Revenue per customer trend
Works when:
•
Clear unit of value (transaction, message, call)
•
Customers prefer pay-as-you-go
•
Usage correlates with value received
•
You have efficient operations at scale
Marketplace
Connect buyers and sellers, taking a cut of transactions.
Revenue: Transaction fees, listing fees, or subscriptions
Examples: Airbnb, Uber, Upwork, Etsy
Characteristics:
•
Chicken-and-egg problem (need both sides)
•
Network effects create moats
•
Winner-take-most dynamics
•
Regulatory considerations
Key metrics:
•
Gross Merchandise Value (GMV)
•
Take rate
•
Liquidity (transactions per listing)
•
Supply and demand balance
Works when:
•
Fragmented supply and demand
•
Value in aggregation
•
Trust problems you can solve
•
Large enough market for network effects
Freemium
Basic product is free, premium features require payment.
Revenue: Subscriptions from converted free users
Examples: Dropbox, Spotify, Zoom, Slack
Characteristics:
•
Large top of funnel
•
Low conversion rates (2-5% typical)
•
Free users have cost
•
Requires viral/organic growth
Key metrics:
•
Free to paid conversion rate
•
Time to conversion
•
Cost to serve free users
•
Viral coefficient
Works when:
•
Product has viral mechanics
•
Clear upgrade triggers
•
Low marginal cost for free users
•
Large market to penetrate
Open Core
Core product is open source, enterprise features are paid.
Revenue: Enterprise licenses, support, hosted versions
Examples: GitLab, MongoDB, Elastic, HashiCorp
Characteristics:
•
Community builds awareness and adoption
•
Enterprise features drive revenue
•
Balance between open and paid features
•
Competition from your own project
Key metrics:
•
Community size and engagement
•
Enterprise conversion
•
Revenue per enterprise customer
•
Contribution velocity
Works when:
•
Developer/technical audience
•
Network effects from adoption
•
Clear enterprise needs beyond open source
•
Community provides value (contributions, bug reports)
Advertising
Product is free, revenue comes from advertisers.
Revenue: CPM, CPC, or sponsorships
Examples: Google, Facebook, Twitter
Characteristics:
•
Need massive scale
•
User experience vs. revenue tension
•
Privacy and data concerns
•
Dependent on ad market conditions
Key metrics:
•
Active users
•
Time on platform
•
Ad inventory (impressions)
•
Revenue per user (ARPU)
Works when:
•
Massive user base possible
•
High engagement/attention
•
Data enables targeting
•
Unlikely for most startups
Choosing Your Model
Match Model to Market
Ask:
•
How do customers currently pay for similar value?
•
What’s the natural unit of value?
•
How frequently do customers receive value?
•
What’s the willingness to pay?
A developer tool might work as usage-based. A team collaboration tool fits SaaS. A local services company fits marketplace.
Match Model to Your Strengths
Different models require different capabilities:
SaaS: Requires customer success, retention focus, ongoing development
Transactional: Requires operational excellence, low costs at scale
Marketplace: Requires growth skills on both sides, community building
Freemium: Requires virality, large potential market, capital to fund free users
Match Model to Funding Reality
Some models require more capital:
Capital efficient: SaaS (if efficient acquisition), transactional
Capital intensive: Marketplace (chicken-and-egg), freemium (funding free users), advertising (need scale before revenue)
If bootstrapping, choose capital-efficient models.
Hybrid Models
Many successful companies combine models:
SaaS + Usage: Base subscription plus usage fees (Vercel, PlanetScale)
Freemium + Enterprise: Free tier, paid pro, enterprise contracts (Figma, Notion)
Marketplace + SaaS: Platform fee plus subscription for sellers (Shopify)
Open Core + SaaS: Open source core, cloud hosted paid version (GitLab, Supabase)
Model Evolution
Your model can evolve:
Slack: Started free/freemium → Added enterprise contracts → Now primarily enterprise revenue
AWS: Usage-based → Added reserved capacity (annual commitments) → Hybrid
Adobe: Perpetual licenses → SaaS subscriptions
Don’t feel locked into your initial model. But changes have implications—plan carefully.
Business Model Red Flags
No clear revenue model: “We’ll figure out monetization later” often means never.
Doesn’t match customer behavior: Subscriptions for infrequent use, per-transaction for daily use.
Unsustainable economics: Negative unit economics that don’t improve with scale.
Requires unrealistic scale: Advertising model that needs Facebook-level users.
Competition commoditizes: Model requires differentiation that doesn’t exist.
Validating Your Model
Before committing:
1.
Talk to potential customers about pricing and willingness to pay
2.
Study how they pay today for similar value
3.
Calculate unit economics at different price points
4.
Test with early customers before formalizing
5.
Watch for signals that model isn’t working (churn, complaints, low conversion)
Key Takeaways
•
SaaS works for ongoing value with predictable pricing needs
•
Usage-based works when usage correlates with value received
•
Marketplace works for fragmented supply and demand with aggregation value
•
Freemium works with viral mechanics and low marginal costs
•
Open core works for developer tools with enterprise upgrade path
•
Match your model to market behavior, your strengths, and funding reality
•
Models can evolve—don’t feel permanently locked in
•
Validate with customers before committing
Put your context to work.
Keep your project’s decisions, next steps and proof together as your agents work.
Explore the agent workflow →