cursor.yml

company: "Cursor"
slug: "cursor"
url: "cursor.com"
sub_industry: "tooling"
gtm: "PLG"
verified_at: "2026-06-22"
metrics:
  - field: "arr"
    value: "$1B+"
    as_of: "2025-11"
    tier: "Verified"
    source_type: "press"
    source_url: "https://cursor.com/blog/series-d"
    attribution: "Cursor official blog (Series D announcement)"
    note: "\"We've also crossed $1B in annualized revenue.\" Announced Nov 13, 2025 alongside Series D."
  - field: "funding_round"
    value: "Series D, $2300M"
    as_of: "2025-11"
    tier: "Verified"
    source_type: "press"
    source_url: "https://cursor.com/blog/series-d"
    attribution: "Cursor official blog (Series D announcement)"
    note: "\"our Series D of $2.3B at a $29.3B post-money valuation\"; co-led by Accel, Coatue, with NVIDIA, Google, Thrive, a16z. Matches row's Series D and 2300 amount."
  - field: "nrr"
    value: "250%"
    as_of: "2024-03"
    tier: "Reported"
    source_type: "journalism"
    source_url: "https://x.com/theinformation/status/1905990843623768279"
    attribution: "The Information (official X account)"
    note: "Net revenue retention for the March 2024 customer cohort; reported by The Information, not company-disclosed."
  - field: "ttfv"
    value: "instant"
    as_of: "2025"
    tier: "Estimated"
    source_type: "estimate"
    source_url: ""
    attribution: "EM internal estimate / product-marketing descriptions"
    note: "No primary or credible-journalism source quantifies time-to-first-value. Plausible given autocomplete works immediately on install, but unsourced."
  - field: "retention"
    value: "95-97% monthly"
    as_of: "2024"
    tier: "Estimated"
    source_type: "aggregator"
    source_url: ""
    attribution: "DevGraphiq stat roundup (aggregator)"
    note: "Implied from a claimed 3-5% monthly churn; traces only to DevGraphiq aggregator, no primary or named-journalism source."
  - field: "free_to_paid"
    value: "35%"
    as_of: "2024"
    tier: "Estimated"
    source_type: "aggregator"
    source_url: ""
    attribution: "DevGraphiq / stat-roundup aggregators"
    note: "35-36% free-to-paid repeated across aggregator stat pages (DevGraphiq, shipper.now); no primary company source or named journalism retrieved."
  - field: "organic"
    value: "~100% ($0 marketing)"
    as_of: "2025"
    tier: "Estimated"
    source_type: "estimate"
    source_url: ""
    attribution: "EM characterization; $0 marketing widely repeated in secondary coverage"
    note: "Multiple secondary sources state the team grew with no paid marketing / no outbound sales org; the specific ~100% organic figure is an EM characterization, no primary or single named-journalism URL with the figure retrieved."
plays:
  - title: "Fork the incumbent so switching to you costs nothing"
    summary: "Cursor forked VS Code, inheriting the entire extension and theme ecosystem. On first launch it detects an existing VS Code install and offers a one-click import of extensions, themes, keybindings, and all settings. Switching cost a developer almost nothing, so Cursor competed on its AI features rather than fight years of muscle memory."
    goal: "Activation"
    learning: "Against a sticky incumbent, build on its forkable foundation and ship a one-click import of the user's existing extensions, themes, and keybindings on first launch. Trial then costs them zero setup, freeing you to compete on your new value rather than their accumulated muscle memory."
    when_it_works: "Works when switching pain comes mostly from setup and muscle memory, with little lock-in from data or contracts."
    metric_moved: "ttfv"
    result: "Developers keep their full VS Code environment on switch; the fork let Cursor inherit the VS Code ecosystem while owning the AI layer, a key enabler of its zero-marketing, word-of-mouth growth to $1B+ ARR."
    tradeoff: "Forking VS Code chains you to Microsoft's upstream: you carry the permanent cost of rebasing your AI layer onto their releases and you depend on a competitor's terms for the proprietary extension marketplace, which they can restrict against forks at any time."
    tier: "Verified"
    source_type: "filing"
    source_url: "https://cursor.com/docs/configuration/migrations/vscode"
    attribution: "Cursor official docs: \"VS Code Import\" one-click import of extensions, themes, settings, and keybindings"
    tags:
      - "onboarding-redesign"
      - "bottom-up-plg"
      - "open-source"
      - "default-on"
      - "migration"
  - title: "Win the first session with one on-by-default wow"
    summary: "Cursor's wedge was Tab, a custom model that predicts your next edit (not just the next token) and fires the instant you type, no setup or prompting. By Cursor's own account Tab \"opened the first era of AI-assisted coding\" and gave \"significant leverage for nearly two years.\" The on-by-default wow turned trials into daily habits and made developers evangelize unprompted."
    goal: "Activation"
    learning: "Engineer a single wedge feature that shows a visible payoff within seconds of install, ship it on by default with zero configuration, and let that one moment drive trial-to-habit and word-of-mouth. For Cursor it was Tab predicting your next edit the instant you typed."
    when_it_works: "Works when a genuinely surprising result is reachable without setup, data, or a learning curve to trigger it."
    metric_moved: "growth"
    result: "Tab was the PMF trigger and the primary feature through early 2025 (2.5x more users than agents as of March 2025); it anchored Cursor's rise to the fastest SaaS company ever to $100M ARR (12 months)."
    tradeoff: "By Cursor's own account Tab gave leverage for only \"nearly two years,\" so a single copyable wedge feature buys a finite head start that erodes the moment rivals replicate the trick and the surprise becomes table stakes."
    tier: "Verified"
    source_type: "press"
    source_url: "https://cursor.com/blog/third-era"
    attribution: "Cursor official blog, \"The third era of AI software development\""
    tags:
      - "default-on"
      - "developer-advocacy"
      - "word-of-mouth"
      - "wedge"
  - title: "Design the free tier as the first rung of a team expansion ladder"
    summary: "Cursor ran a bottom-up loop with no outbound sales for years: a free tier (2,000 completions/month) seeds personal use, the developer hits the contrast with single-line Copilot and upgrades to $20/mo Pro, mentions it in standup so teammates try it, the manager sees faster shipping, and the org buys a Business plan. The same engineers surface it at conferences, restarting the loop. A 40-60 person team rode this with effectively $0 paid marketing."
    goal: "Acquisition"
    track: "Word-of-mouth & outbound"
    learning: "Make individual value obvious enough that solo users upgrade on their own, then build visible output and shared workflows so paid individuals pull teammates in and managers buy the team plan. That bottom-up loop replaces an outbound sales motion entirely."
    when_it_works: "Works when individual value is felt before team value and adoption is visible to peers."
    metric_moved: "organic"
    result: "Reported $0 marketing spend to $100M ARR and on to $500M ARR with no traditional sales/marketing; 35% free-to-paid conversion (roughly 10x the 1-4% SaaS norm) feeds the loop."
    tradeoff: "A generous free tier hands real model-inference COGS to users who may never pay, and the whole loop stalls wherever the buyer is top-down and individual usage stays invisible inside the org, so you surrender control of the sales motion to whether peers happen to notice each other."
    tier: "Reported"
    source_type: "journalism"
    source_url: "https://www.builderlab.ai/p/growth-machines-the-cursor-story"
    attribution: "\"Growth Machines: The Cursor Story,\" BuilderLab (Pranav Pathak); corroborated by digidai analysis of Michael Truell / Cursor growth (Nov 2025)"
    tags:
      - "word-of-mouth"
      - "bottom-up-plg"
      - "free-tier"
      - "community-led"
      - "land-and-expand"