AI summary by TwinMind
How To Price For B2B | Startup School
Y Combinator
17m 45s
Introduction & Common Pricing Problem
- Tom, Y Combinator partner, addresses most common founder question: pricing
- Scenario: founder working outbound sales, successful call, champion interested, asked for pricing, froze
- Problem: founders without big company experience lack calibration for B2B software prices
- Common mistake: referencing consumer software prices ($19-49/month GitHub, ChatGPT) leading to ludicrously low pricing
- Psychological barrier: asking for tens/hundreds of thousands feels uncomfortable, can't say with straight face
Core Pricing Framework: Three Elements
1. Value Equation (Most Important - 80-90% of pricing)
- Sit with champion to document expected value delivery
- Three value types: cost saving, time saving, revenue increase
- Write down step-by-step with champion challenging assumptions
- Tool for champion to justify purchase to boss/CFO
Example Calculation:
- Customer service tool for company with 100 support agents
- Each agent: $50K salary + $50K additional costs = $100K fully loaded cost
- Total customer service cost: $10M annually
- AI tool eliminates 20% queries/time = $2M potential savings
- Pricing: 25-50% of value delivered
- Customer keeps ~2/3, vendor keeps ~1/3
- Price: ~$700K (customer saves $1.3M net)
2. Cost Considerations
- Never start with cost-plus pricing - always underprices software
- Cost only serves as floor, not starting point
- Target 80-90% gross margins for software
- Example: $700K contract value, $200K costs (OpenAI/AWS) = good business
- Bad scenario: $150K contract value, $200K costs = unsustainable
- Credit warning: treat AWS/Microsoft/OpenAI credits as cash cost, don't assume unlimited forever
- Pricing below cost occasionally acceptable for market share grab but "really really risky"
- LLM cost reductions may justify temporary lower margins but still risky
3. Competition
- Avoid price wars - race to the bottom
- Don't engage head-to-head bidding for commodity products
- Differentiate based on functionality/value, not price
- Airline industry example: 2.7% net profit margin due to commodity nature
- Solution: make product non-comparable through differentiation
Pricing Structure & Strategy
Industry Research
- Ask champion about payment preferences for similar software
- Understand industry norms: monthly flat fee, per-seat, usage bands, credits
- Mirror familiar pricing structures
- People wary of uncapped usage pricing - put caps
- Keep pricing simple - complexity kills sales
Recurring vs Usage-Based Revenue
- Prefer committed recurring revenue (MRR/ARR) over usage-based
- Economic downturn protection until contract renewal
- Usage-based revenue can "fall off cliff" in bad months
- Technique: start usage-based, then convert to minimum monthly commitment with volume discounts
- Example: $15K/month average usage → $12K/month flat fee for 12-month commitment
Signing Authority Optimization
- Ask champion's personal signing authority limit
- Example: $15K authority → price pilot at $14,999 to avoid additional approvals
Enterprise vs Self-Service Pricing
Website Pricing Decision
- Strong internet opinions: developers want visible prices, hate "contact sales"
- Enterprise value equations differ per customer
- Random website pricing leaves money on table
- Overprices for low-value customers (lose them)
- Underprices for high-value customers (lost revenue)
Typical Structure
- 1-2 cheaper self-service plans (individual, small team)
- Basic functionality included, enterprise features excluded
- Enterprise-gated features: SOC 2 audits, SSO, audit logs, compliance reports, data geography
- Enterprise pricing can be 10x higher for compliance/legal features
Sales Team Economics
5:1 Rule
- 5:1 ratio between new signed ARR and total salesperson compensation
- Example: $100K total comp → expect $500K new ARR annually
Contract Size Impact on Sales Model
- $500K single contracts: whale hunting, 4-6 deals simultaneously, one every couple months
- $25K contracts ($2K monthly): ~2 closes per month, 20 annually - still account executive model
- $1K annual ($83 monthly): 42 deals monthly, almost 2 per working day - call center/inside sales model
Trials vs Pilots
Free Trial Problems
- Long free trials counterproductive
- Customer not bought in without payment
- Keep pilots short: 2-4 weeks maximum
- Require clear success criteria from value equation
Better Approach
- Annual contract with 30-60 day money-back guarantee
- Opt-out at start if doesn't deliver
- Counts as recurring revenue immediately
- Default becomes recurring contract
Startup Positioning
Don't Fake Size
- Don't add fake employees to website/LinkedIn
- Don't pretend to be bigger company
- Play to startup strengths: founder phone numbers, 24/7 support
- Advantage over Salesforce/Oracle: personal attention
Pricing Experimentation
When Uncertain
- Pick number similar to comparable software
- Increase by 50% for each new customer pitch
- Example progression: $10K → $15K → $22K
- Stop when losing >25% of deals solely on price
- Don't need to win every deal
- If closing every deal → definitely underpricing
Long-term Perspective
- Initial 5-10 customers = tiny fraction of 5-year revenue
- More important to start deal flow than optimize early pricing
- Can increase prices as product improves
- Add new modules behind paywall for upselling
- Over-optimizing early pricing = big mistake
- First 2-3 sales hardest you'll ever do
Success Metrics from Value Equation
- Provides pilot success metrics to measure
- Example: 20% query reduction target
- Adjust pricing based on actual results (15% vs 25% savings)
- Metrics prove value equation validity
