Case Study Comparison: 19 Companies, Every Framework
All 19 case studies mapped against each other. The patterns that emerge reveal what’s universal and what’s context-dependent.
The Master Table
| Company | Founded | Outcome | Category | Funding | PMF Time | Moat Type |
|---|---|---|---|---|---|---|
| Airbnb | 2008 | $100B+ IPO | Marketplace | ~$6B | ~2 years | Two-sided marketplace |
| Stripe | 2010 | $50B+ private | Infrastructure | ~$2.3B | ~1 year | Developer ecosystem |
| Slack | 2013 | $27.7B acquired | SaaS | ~$1.4B | Immediate | Workflow embedding |
| Shopify | 2006 | $100B+ IPO | Platform | ~$122M | ~2 years | Platform ecosystem |
| WeWork | 2010 | Bankruptcy | Real estate | ~$12B+ | Never | None |
| Cursor | 2022 | $29B private | AI tool | ~$3.2B | ~6 months | Speed + UX + data |
| Stitch Fix | 2011 | $730M IPO | Data + retail | ~$42M | ~1-2 years | Data network effects |
| Gumroad | 2011 | $10M+ ARR (indie) | Creator tools | ~$8M (returned) | ~1 year | Creator lock-in |
| Midjourney | 2022 | $500M+ ARR | AI generation | $0 (bootstrapped) | Immediate | Community + brand |
| Basecamp | 2004 | $100M+ (private) | SaaS | $0 (bootstrapped) | ~1 year | Philosophy as brand |
| Levels | 2019 | $3M+ ARR (solo) | Indie SaaS | $0 (bootstrapped) | Months | Speed + shipping pace |
| Linear | 2019 | $1.25B private | Dev tools | ~$134M | ~14 months | Craft + opinionated UX |
| Lovable | 2023 | $6.6B private | AI builder | ~$553M | ~1 month | Brand + speed |
| Kit | 2013 | $43M ARR | Creator tools | $0 (bootstrapped) | ~2.5 years | Community + niche |
| Notion | 2013 | $11B private | Workspace | ~$343M | ~5 years | Blocks architecture |
| Fast | 2019 | Shutdown | Payments | ~$125M | Never | None |
| Perplexity | 2022 | $20B+ private | AI search | ~$1.7B | ~6 months | Retrieval + trust |
| Case studies pending |
Success rate: 17 of 19 survived (89%). 2 failures (WeWork, Fast). But survivorship bias applies — we chose these companies because they’re instructive.
The Four Eras
Era 1: Web 1.0 / Pre-Mobile (2004-2010)
Companies: Basecamp (2004), Shopify (2006), Airbnb (2008), Stripe (2010), WeWork (2010)
Pattern: Longer timelines to PMF (1-2+ years). Physical-world integration common (Airbnb=stays, Shopify=shipping, WeWork=offices). Bootstrapping was default for some (Basecamp, Shopify early). Capital efficiency varied wildly (Shopify $122M → $100B+; WeWork $12B → $0).
Era 2: Mobile / SaaS (2011-2018)
Companies: Stitch Fix (2011), Gumroad (2011), Slack (2013), Kit (2013), Notion (2013)
Pattern: PLG emerged as dominant growth model. Mobile distribution created new channels. SaaS pricing models matured. Notion’s 5-year PMF journey is the outlier — most companies either found PMF faster or died.
Era 3: Cloud / Developer Tools (2019-2021)
Companies: Linear (2019), Levels (2019), Fast (2019)
Pattern: Developer-first distribution. Community as moat. Solo founder viability emerging (Levels). But also ZIRP excess (Fast: $125M raised, $600K revenue).
Era 4: AI-Native (2022-present)
Companies: Cursor (2022), Midjourney (2022), Perplexity (2022), Lovable (2023)
Pattern: Radically compressed timelines. Cursor: $0 → $1B ARR in <24 months. Lovable: $100M ARR in 8 months. Team sizes 10-50x smaller. Bootstrapping viable at massive scale (Midjourney: $500M+ ARR, zero VC). Model-agnostic strategies emerging (Perplexity uses 5+ models).
Framework Comparisons
ideation: How Ideas Emerge
| Pattern | Companies |
|---|---|
| Scratch own itch | Shopify, Basecamp, Linear, Notion, Perplexity |
| Schlep Blindness | Stripe (payments tedium) |
| Feature pivot | Slack (game → chat), Gumroad (side project → main product) |
| Domain clash | Stitch Fix (data science × fashion), Lovable (open source → product) |
| Organic / personal pain | Airbnb, Cursor, Kit, Levels |
| Market observation | WeWork (only failure in this category) |
| Frontier bet | Perplexity (AI search), Midjourney (AI generation), Cursor (AI coding) |
19-company pattern: 18 of 19 started from personal experience or frontier technical insight. WeWork is the sole market-observation-only founder — and the biggest failure.
product-market-fit: Time to PMF
| Speed | Companies | What Explains It |
|---|---|---|
| Immediate | Slack, Midjourney | Instant pull; users couldn’t stop |
| <6 months | Cursor, Lovable, Perplexity, Levels | AI-era speed; rapid iteration |
| 1-2 years | Airbnb, Stripe, Shopify, Stitch Fix, Basecamp, Kit (eventually), Gumroad | Standard SaaS/marketplace timeline |
| 2-5 years | Notion (4 rebuilds), Linear (14-month beta) | Patience + craft approach |
| Never | WeWork, Fast | Demand without viable economics = fake PMF |
Key insight: AI-era companies find PMF in months, not years. But patience works too — Notion rebuilt 4 times over 5 years and reached $600M ARR.
do-things-that-dont-scale: Early Tactics
| Tactic Type | Companies |
|---|---|
| Manual installation | Stripe (Collison installation), Kit (concierge migration) |
| Door-to-door | Airbnb (photography), Kit (emailing bloggers) |
| Handpicked beta | Linear (10/week for 14 months), Superhuman-style |
| Personal delivery | Stitch Fix (carrying clothes to homes) |
| Build in public | Levels (shipped 100+ projects publicly), Lovable (Anton on X) |
| Kyoto rebuild | Notion (moved to Japan, coded in underwear for a year) |
| Discord community | Midjourney (entire product lived in Discord) |
19-company pattern: All 19 did unscalable things early — including both failures. Unscalable tactics are necessary but not sufficient.
competitive-strategy: Zero-to-One?
| Created new category | Competed in existing market |
|---|---|
| Airbnb (home stays) | WeWork (real estate) |
| Stripe (developer payments) | Fast (checkout — already solved) |
| Slack (team chat UX) | |
| Shopify (merchant empowerment) | |
| Cursor (AI-native editor) | |
| Perplexity (answer engine) | |
| Midjourney (AI art) | |
| Lovable (AI app builder) | |
| Linear (opinionated dev tools) | |
| Notion (blocks-based workspace) |
17 of 19 created new categories or fundamentally new approaches. Both failures competed in existing markets without genuine differentiation.
moats: What Creates Defensibility
| Moat Type | Companies | Durability |
|---|---|---|
| Two-sided marketplace | Airbnb, Shopify | Very durable |
| Developer ecosystem | Stripe, Shopify, Cursor | Very durable |
| Workflow/switching costs | Slack, Linear, Notion | Durable |
| Data network effects | Stitch Fix, Cursor, Perplexity | Growing |
| Community | Midjourney, Kit, Notion | Moderate-durable |
| Craft/brand | Linear, Basecamp, Notion | Moderate |
| Speed of execution | Lovable, Levels, Cursor | Fragile |
| None | WeWork, Fast | — (both failed) |
Pattern: Companies with 2+ moat types survived turbulence. Companies with zero moats failed. Speed is a fragile moat — it buys time to build durable ones.
fundraising: Capital Efficiency
| Tier | Company | Raised | Outcome | Efficiency |
|---|---|---|---|---|
| Outstanding | Midjourney | $0 | $500M+ ARR | ∞ |
| Outstanding | Basecamp | $0 | $100M+ private, 25+ years | ∞ |
| Outstanding | Levels | $0 | $3M+ ARR solo | ∞ |
| Outstanding | Stitch Fix | $42M | Profitable IPO | Excellent |
| Outstanding | Kit | $0 | $43M ARR bootstrapped | ∞ |
| Excellent | Shopify | $122M | $100B+ | $819 return per $1 |
| Excellent | Linear | $134M | $1.25B, profitable, negative burn | Excellent |
| Good | Stripe | $2.3B | $50B+ | Good |
| Good | Slack | $1.4B | $27.7B | Good |
| Good | Airbnb | $6B | $100B+ | Good |
| Good | Notion | $343M | $11B, $600M ARR | Good |
| TBD | Cursor | $3.2B | $29B, $1B ARR | TBD |
| TBD | Perplexity | $1.7B | $20B+, $200M+ ARR | TBD |
| TBD | Lovable | $553M | $6.6B, $400M ARR | TBD |
| Poor | Gumroad | ~$8M (investors sold for $1) | $10M+ ARR | Restructured |
| Catastrophic | WeWork | $12B+ | Bankruptcy | -100% |
| Catastrophic | Fast | $125M | $600K revenue, shutdown | -100% |
5 of 19 bootstrapped to significant outcomes without VC. More money ≠ better outcome — it just raises the stakes.
Success vs. Failure: The 2 Failures
| Factor | 17 Successes | WeWork | Fast |
|---|---|---|---|
| Unit economics | Worked or trending positive | Never worked ($1.9B loss on $1.8B revenue) | Never worked ($166 per $1 earned) |
| Structural moat | At least one | Zero (0 of Thiel’s 4) | Zero |
| Honest metrics | Real numbers | ”Community Adjusted EBITDA” | No public metrics until exposé |
| Zero to one | Created new category | Competed in existing market | Solved an already-solved problem |
| Capital discipline | Raised what they needed or bootstrapped | Raised everything available | Hired 480 on $600K revenue |
| Founder pattern | First-time or proven repeat | Unchecked charisma | Prior business collapses |
| Time from exposure to death | N/A | Months (IPO pulled) | 6 days |
The single best predictor of failure: zero structural moats. Both failures scored 0 on the moats framework.
The AI-Era Shift (Expanded)
| Metric | Pre-AI Best | AI-Era Companies |
|---|---|---|
| Time to $100M ARR | Slack: ~3 years | Lovable: 8 months |
| Time to $1B ARR | Slack: ~6 years | Cursor: <24 months |
| Employees at $100M ARR | Slack: hundreds | Lovable: 45 |
| ARR per employee | Linear: $700K | Lovable: $2.77M |
| VC required? | Usually yes | Midjourney: $500M+ ARR, $0 VC |
| Solo founder viable? | Rare (Stitch Fix) | Common (Levels: $3M+ ARR, 0 employees) |
| Distribution | Paid + content + sales | Product virality + community + build-in-public |
What changed: team size (10-50x smaller), speed (3-5x faster), bootstrapping viability (profitable from day one possible). What didn’t change: PMF still required, unit economics still matter, moats still determine survival, founder determination still wins.
The Counterintuitive Findings
-
Patience and speed both work. Notion took 6 years to $3M ARR then exploded. Lovable hit $100M in 8 months. There’s no single right timeline — but you must be honest about whether you’re being patient or just slow.
-
Bootstrapping produces higher survival rates. 5 of 5 bootstrapped companies survived (100%). 12 of 14 VC-funded survived (86%). Small sample, but the pattern is real — external capital adds pressure that kills marginal companies.
-
The founder’s past predicts the company’s future. Holland (Fast) had prior business collapses. Neumann (WeWork) had unchecked charisma. Every successful founder had either genuine domain expertise, technical depth, or both.
-
Category creation is nearly mandatory. 17 of 17 survivors created or fundamentally redefined their category. Both failures competed in existing markets.
-
The “right” amount of funding varies 12,000x. From $0 (Midjourney) to $12B (WeWork). The determining factor isn’t the amount — it’s whether the business underneath justifies it.
See Also
- case-study-airbnb | case-study-stripe | case-study-slack | case-study-shopify
- case-study-wework | case-study-cursor | case-study-stitch-fix
- case-study-gumroad | case-study-midjourney | case-study-basecamp | case-study-levels
- case-study-linear | case-study-lovable | case-study-kit | case-study-notion
- case-study-fast | case-study-perplexity
- the-startup-lifecycle | where-the-experts-disagree | start-here