Skip to content

samuelcastro/startup-skills / idea-validation

idea-validation

Verified

Guide founders through validating startup ideas before building. Use when a founder says "I have an idea", "help me validate", "is this a good idea", "should I build this", "customer discovery", or asks about product-market fit, market sizing, or customer interviews. Covers problem discovery, customer interviews, market validation, and go/no-go decisions.

FileKindSize
SKILL.mdskill md5.3 KB
references/case-studies.mdmarkdown10.6 KB
references/frameworks.mdmarkdown4.9 KB
references/interview-guide.mdmarkdown5.7 KB
references/traps.mdmarkdown6.4 KB

references/case-studies.md

View on GitHub
# Founder Case Studies

Eight stories of how successful founders validated their ideas. Source: Todd Jackson, First Round Capital.

## Three Paths to Finding Ideas

| Approach | Best For | Examples |
|----------|----------|----------|
| **Market First** | B2B, enterprise | Vanta, Cocoon |
| **Experience First** | Consumer, marketplace | Good Dog, Snackpass, Rec Room |
| **Problem First** | Dev tools, infra, B2B | LaunchDarkly, Pinwheel, Flexport |

---

## 1. Vanta — Christina Cacioppo

**Category**: B2B SaaS (Security/Compliance)
**Approach**: Market First

### Finding the Idea
- Started on a couch brainstorming for months. Her advice: "Don't do this."
- Pivoted to investigating security space out of interest
- Interviewed founder friends and security leaders
- Asked: "What's the best and worst part of your day?"
- Heard security was blocker to moving upmarket and unlocking revenue

### Validation Method
- Built gap assessment (like SOC 2 consultant would create)
- Created zero-code MVP: a spreadsheet
- Shared with friends, former coworkers, networks

### Lightbulb Moment
> People who Christina hadn't even spoken to started reaching out directly to ask about using Vanta.

### Traction
- Joined YC, first full audit Nov 2018
- 30 customers by end of year
- Today: 2,800+ customers, $55M raised

### Key Lesson
> "Do all the product development best practices. Nobody wants to do them because they're less fulfilling than building. But it's so easy to change a spreadsheet or mockup; it's much harder to change code."

---

## 2. Cocoon — Mahima Chawla

**Category**: B2B SaaS (HR/Leave Management)
**Approach**: Market First

### Finding the Idea
- Explored pivotal life moments, particularly becoming first-time parents
- Interviewed dozens of parents
- Initially open-ended on financial planning
- Conversations consistently steered to problematic topic of leave

> "In other areas, we were searching for the problem, whereas with leave it was obvious."

### Validation Method
- Put ad on Craigslist to talk to broader audience (hourly workers)
- Heard nightmare stories: not getting paid, 10 hours on hold for benefits
- Validated employee side, then validated employer side separately
- **Designed UI flow in Figma**, shared with interviewees

### Lightbulb Moment
> Signed first two customers—Carta and Benchling—off Figma prototypes alone.

### Traction
- Raised seed Dec 2020
- Inbound leads from word of mouth
- $26M raised, rapid customer growth

### Key Lesson
> "Go in with a very open mind. Don't try to make something a problem, because if you do you're going to find the evidence to support it."

---

## 3. Good Dog — Josh Wais & Lauren McDevitt

**Category**: Consumer Marketplace
**Approach**: Experience First

### Finding the Idea
- Both experienced challenges finding a dog personally
- Saw opportunity: marketplace connecting reputable breeders with consumers
- Mission felt more compelling than anything before

### Validation Method
- Used "sacrificial concepts" (low-fidelity printouts) to test solutions
- Asked friends/colleagues for anyone who recently got a dog
- Compensated interview participants to reduce bias
- **20+ dog owner interviews** asking:
  - How did you decide you wanted a dog?
  - What was most painful part of the process?
  - Which part was most exciting?

### Lightbulb Moment
> On consumer side: hearing people gave up on their dog search altogether due to pain.
> On breeder side: discovering that tapping into breeders' pride and passion motivated them to engage.

### Traction
- Worked 6 months nights/weekends cold-calling breeders
- Launched with 300 breeders March 2019
- 20K+ breeders by end of 2021
- Raised Series B

### Key Lesson
> "Don't ignore your gut. Believing it should exist and having information to support that is really critical. Everything after that is problem solving."

---

## 4. Snackpass — Kevin Tan

**Category**: Consumer/Social Commerce
**Approach**: Experience First

### Finding the Idea
- At Yale, saw pandemonium at popular food spot "G-Heav" every Saturday
- Students struggled to place orders despite Uber Eats existing
- Envisioned order-ahead to skip lines

### Validation Method
- Leveraged relationships with local restaurant owners
- Went door-to-door to convince them to join
- Built website over Thanksgiving break
- Used API to send orders by fax machine
- Launched with 5 restaurants
- Printed flyers, distributed across campus

### Lightbulb Moment
> "It was amazing. It was the first time people signed up for a product I'd made. It was a trickle at first but kept growing."

### Traction
- 90% of Yale students using within 6 months
- Added social layer (points system) for word-of-mouth
- Restaurants calling saying they needed to shut down—couldn't keep up
- Replicated success at Brown (proved not just Yale phenomenon)
- $96.5M raised

### Key Lesson
> "It's natural to think about things in trends, but the jobs-to-be-done framework says people hire products to get something done. That's a more sane way to look at it."

---

## 5. Rec Room — Nick Fajt

**Category**: Consumer/Gaming
**Approach**: Experience First

### Finding the Idea
- Team at Microsoft working on HoloLens
- Realized app model was wrong—tiles didn't work together
- Wanted social substrate stitching experiences together
- Started by building a game (not platform) with 6 people and no money

### Validation Method
- Built first game in 90 days
- Launched quietly on SteamVR
- Set up in WeWork lobby, asked passersby to try app
- Posted on Reddit for testers
- Gave demos to investors from within the game

### Lightbulb Moment
> "We were very surprised. It struck a chord somehow and had a weird, buggy charm to it."

Users invented activities team never built: plays, murder mysteries, hide-and-seek.

### Traction
- Launched free (only free content on SteamVR at time)
- Added Oculus, more games
- Pivoted to UGC and cross-platform despite backlash
- 3M monthly active VR users today
- $294M raised

### Key Lesson
> "The absolute value of emotion you feel on any given day is way higher early in the founder journey. You can go from feeling on top of the world in the morning to feeling like the company will die next week."

---

## 6. LaunchDarkly — Edith Harbaugh

**Category**: B2B/Dev Tools
**Approach**: Problem First

### Finding the Idea
- Saw same problem repeatedly as engineer: releases pushed to everyone, goes horribly wrong
- Knew large companies had internal solutions (Dropbox Gandalf, Facebook Gatekeeper)
- Deep conviction feature flagging could unlock new way to develop software

### Validation Method
- Partnered with co-founder who'd seen problem at Atlassian
- First users were old engineering buddies
- Many admitted they tried it just as favor to Edith
- **Started blogging about feature flagging for SEO/inbound leads**
- Interviewed leaders at Facebook, Netflix for content

### Lightbulb Moment
> On InVision support call, suddenly a dozen employee faces popped up on Zoom. People were invested enough to care about asking questions.
> Later: Signed first six-figure contract when VCs said it was "$5/month dev tool."

### Traction
- First 8 months: only 1 paying customer per month
- 2016: First six-figure contract
- Revenue became their Series A ("Revenue is its own series A")
- Today: 3,000 customers, 22 of Fortune 100, $3B valuation

### Key Lesson
> "Build something you're excited and passionate about but with a market that is more than just you. If you're not excited about the work, it shows."

---

## 7. Pinwheel — Kurtis Lin

**Category**: B2B/Fintech Infrastructure
**Approach**: Problem First

### Finding the Idea
- Started with HSA idea (lived the problem)
- Kept hitting integration challenges with payroll systems
- Wondered: "If integrations are this big a problem for us, maybe for others too"
- Built lead list of ICP, started conversations

> "The moment we mentioned how difficult it was to access income and employment data, people's eyes would start lighting up."

### Validation Method
- Created framework: 32 attributes to rate/stack-rank ideas
- Had 20 ideas before choosing Pinwheel
- Developed 3 hypotheses, pitched all to prospects
- Switched pitch order to see what resonated
- **People always leaned forward on direct deposit part**
- Surveyed broader population to validate implementation

### Lightbulb Moment
> When pitching, prospective customers would interrupt: "Wait, you can do WHAT?"

### Traction
- Came out of stealth June 2020
- Expected 20 inbound leads from TechCrunch article
- Got 133 in first 72 hours (Chase, Wells Fargo, PayPal, Square, etc.)
- $77M raised, 100+ team

### Key Lesson
> "If you're not decreasing cost or increasing revenue, nobody is going to buy your product."

---

## 8. Flexport — Ryan Petersen

**Category**: B2B/Logistics
**Approach**: Problem First

### Finding the Idea
- Ran import business with brother, felt pains of customs/shipping
- Saw clearing customs was challenging for non-experts
- 2008: Started laying out thoughts in Excel
- Envisioned "TurboTax for customs"

### Validation Method
- **Used Photoshop to create fake app screenshots**
- Put on website to see if people would sign up
- Used SEO + Google ads with small budget
- **300+ companies signed up for fake product**
- Conducted phone interviews with sign-ups and friends

Asked:
- Who do you use for customs today?
- What are you paying?
- What are your pain points?
- Magic wand—what problem would you make go away?

### Lightbulb Moment
> Saudi Aramco signed up for the mock product. "I thought I was going to get Amazon merchants and small companies, but then the biggest oil producer in the world signed up. I realized this could be much bigger."

### Traction
- 2010-2013: Built v1 (licensing, software, compliance)
- Discovered customs-only was incomplete—companies wanted customs + freight bundled
- Pivoted to add freight earlier than planned
- First big customer: major watchmaker (200 containers)
- Grew 4-5x per year ($2M → $10M → $50M)
- Today: $5B revenue, 4th largest US freight forwarder
- $2.2B raised

### Key Lesson
> "A lot of people think if they can't raise venture capital, they shouldn't do a startup. But pick something you can do, get started, and try to get traction before raising."

---

## Summary: Validation Tactics Used

| Founder | Key Tactic |
|---------|-----------|
| Christina (Vanta) | Zero-code spreadsheet MVP |
| Mahima (Cocoon) | Figma prototypes, sold before building |
| Josh/Lauren (Good Dog) | Sacrificial concepts, paid interviews |
| Kevin (Snackpass) | Built over break, fax machine orders |
| Nick (Rec Room) | WeWork lobby demos, Reddit posts |
| Edith (LaunchDarkly) | SEO blogging, leveraged network |
| Kurtis (Pinwheel) | 32-attribute scoring, pitch testing |
| Ryan (Flexport) | Fake screenshots, landing page signups |
slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack."Stars: 198kUpdated Claude CodeCodexVerifiedInstall
ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt.Stars: 194kUpdated Claude CodeCodexVerifiedInstall