AI summary by TwinMind

B2B Startup Metrics | Startup School

Y Combinator

23m 47s

Introduction to Startup Metrics Importance

  • Metrics enable better decision-making for startups
  • Without metrics = "flying airplane with no instruments" - flying blind, not in control
  • Common problem: founders launch on Hacker News/Product Hunt, get hundreds of users but don't know:
    • New vs returning users
    • Daily/weekly active users
    • Churn rates
  • Investors easily distinguish founders who command their metrics vs those who don't
  • Impressive when founders can discuss:
    • Percentage of signups that are DAU/WAU
    • Annual revenue per user

Warning Against Metric Extremes

  • Avoid having 500+ metrics dashboard before launch
  • Don't try to A/B test everything with small user base (few hundred/thousand users)
  • Split test important decisions ($80/year vs $200/year pricing) not trivial ones (button colors)
  • Don't hide behind metrics - still need customer interaction
  • Example: Brian from Airbnb still hosts users in his home

Setting Up Initial Metrics Framework

  • Pick 4-5 key metrics to track accurately (not 30-50)
  • Use straightforward analytics solution:
    • SQL database with simple queries
    • PostHog (Winter 2020 YC company) for SQL analytics
  • Agree on metric definitions and stick with them
  • Avoid internal team conflicts over definitions
  • Example conflict: Marketing claims 2,500 new leads, Sales says they weren't qualified leads
  • Centralized written definitions everyone agrees on

Maintaining Metric Consistency

  • Don't change metric definitions when results disappoint
  • Example: switching from weekly active to monthly active users for better numbers
  • Keep definitions consistent over time to track improvement
  • Inter-company comparisons meaningless due to varying definitions
  • Example: Monzo's weekly active user = financial transaction at least once/week
  • Competitors used different definitions (every 2-8 weeks)

Vanity Metrics vs Core Metrics

  • Early internet: page views, unique visitors (big numbers but not tied to success)
  • Modern vanity metrics:
    • GMV (Gross Merchandise Value) - total dollar value of goods sold
    • Gross Transaction Value for fintech
  • Example: Monzo could report $50 billion transactions/year but revenue very different
  • Revenue should be key metric for most B2B companies

Case Study: Middle East Neo Bank

  • Reported 50% GTV growth every 2 weeks
  • Reality: signing bigger customers with massive cash back/rebates
  • GTV rising but revenue flat for 2 months
  • Founders fooling themselves into thinking company succeeding

Essential B2B Metrics

  • Revenue: Primary metric, don't hide if zero
  • Example: Impressive founder sent 10 monthly investor updates with "big zero" as main metric
  • Burn Rate: Net monthly costs minus revenues (amount bank balance decreases monthly)
  • Runway: Function of burn rate
    • Example: $1M in bank, $100K/month burn = 10 months runway
  • These 3 numbers absolutely crucial for investor updates

Consumer vs B2B Metrics

  • Consumer companies: revenue important but early days focus on:
    • Critical mass/network effects
    • Growing active user base may be more important than revenue initially

Deep Dive: Retention Metrics

  • Retention Rate: If 100 customers sign up in January, how many still paying in Feb/Mar/Apr
  • Measure by cohorts (January cohort, February cohort, etc.)
  • Visualization methods:
    • Heat maps
    • Decay curves over time
    • Stacked cohort visualization (recommended)

Stacked Cohort Visualization Benefits

  • High retention (80-90-100%) = "fat" cohorts over time
  • Builds "layer cake" effect over 2-3 years
  • Example: 18-month graph with 18 cohorts still paying
  • Go Cardless example: Recurring payments company, very sticky customers
  • Team can "go on holiday" and revenue stays consistent
  • Expanding revenue: Customers grow, increasing payments over time

High vs Low Retention Impact

  • High retention: Unstoppable growth, layers building constantly
  • Low retention: "Leaky bucket" scenario
  • Customers churn by month 3-6-9
  • Scrambling to fill bucket as fast as it leaks
  • Natural plateau where working hard just to replace churned customers
  • Very hard to build big business with zero retention

Net Dollar Retention (NDR)

  • B2B SaaS metric focusing on revenue retention
  • Example: AI customer service chatbot
    • January: 10 customers × $10K/month = $100K MRR
    • 12 months later: 2 customers canceled (-$20K), 3 customers upsold to $20K (+$30K)
    • Net result: $110K MRR = 110% NDR
  • Above 100% NDR: Cohorts growing over time
  • Below 100% NDR: Cohorts shrinking, need more new customers

NDR Benchmarks and Reasons

  • Early stage B2B SaaS should target 125-150%+ NDR
  • Mature companies: 110-120% is good
  • Below 100% NDR: Something wrong, focus on fixing vs adding new customers
  • Three reasons early stage should have high NDR:
    1. Underpriced initial product
    2. Adding features/improving product
    3. Getting better at sales/upselling

Deep Dive: Gross Margin

  • Formula: Revenue - Cost of Goods Sold (COGS)
  • COGS: Any cost that varies per customer/incremental customer
  • Software example: AI customer service bot using OpenAI/Anthropic credits
  • Traditional example: Grocery store sandwich = bread + butter + filling costs

Gross Margin Evolution

  • Historical B2B SaaS: 95% gross margins (minimal COGS like AWS bills)
  • Modern software: Lower margins due to AI/operational costs
  • Warning: Free credits don't eliminate costs, just hide them temporarily

Operational Business Challenges

  • Low margin businesses: 5-15% gross margins vs 95% for pure software
  • Examples: grocery delivery, house painting, heat pump installation
  • Solution: Extract software-only version to sell to other companies
  • Higher margins, easier scaling

Zero Interest Rate Era (2010-2021)

  • Blitz scaling: Scaling negative margin businesses with cheap capital
  • Uber example:
    • Negative gross margin initially ($10 service for $9 charge)
    • Subsidized drivers and riders for network effects
    • Burned tens of billions before achieving profitability
  • Other examples: 10-minute grocery delivery, electric scooters

Current Environment and Negative Margins

  • Higher interest rates: Capital more expensive
  • Investors: Very reluctant to fund negative margin businesses
  • Monzo example:
    • Lost £30-40 per customer for first 500K customers
    • Had plan to fix: brought technology in-house, introduced charges, new products
    • Flipped to +£30-40 per customer, now profitable

Final Recommendations

  • Track 4-5 key metrics before launch - don't launch blind
  • Be rigorous: Track right metrics, avoid vanity metrics
  • Clear definitions: Central measurement system, avoid pointless arguments
  • Don't hide behind metrics: Still need customer conversations and product intuition
  • Balance: Metrics + customer feedback + product intuition = vital blend

TwinMind

Get summaries like this for your own meetings

TwinMind takes notes in the background during meetings, lectures, and calls, then writes the summary for you.

Free plan to get started

No bot joins your calls

Phone, desktop & Chrome apps

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome