Chris Dixon at Startup School 2013

Core Thesis

• The "Bad Idea" Paradox: High-potential startups often occupy the "sweet spot" of being ideas that appear bad to outsiders but are actually good, because obvious good ideas are already saturated by academia, governments, and large corporations.
• The "Leftovers" Strategy: Investors and entrepreneurs effectively operate in the business of identifying "leftovers"—opportunities dismissed by mainstream incumbents as unviable or niche.
• Definition of a Secret: A "secret" is defined (per Peter Thiel) as a truth that the speaker believes is true, while the majority of other people do not believe it.

Historical Case Studies of Contrarian Success

• Google (1998):
  • Entered a search market dominated by portals (Yahoo, Lycos, Excite) that viewed search as a loss leader to maximize "stickiness" for ad revenue.
  • Google's core insight was the counter-intuitive strategy of making users leave the site immediately after finding results, rather than keeping them there.
  • Incumbent CEOs rejected Google's technology because it worked "too well," explicitly asking for results that were worse to ensure user retention.
  • A famous anecdote describes a VC sneaking out of a garage meeting to avoid associating with a "bad idea" search engine.
• Airbnb:
  • Initially dismissed as a "hipster" niche activity involving strangers staying on couches in specific neighborhoods (e.g., The Mission, Brooklyn).
  • Perceived as a weird behavior with no path to scaling beyond the travel industry.
  • Evolved into a potential replacement for the hotel industry, validating the initial contrarian view.
• eBay:
  • Launched when the internet was in early commercialization stages without modern payment infrastructure (e.g., PayPal).
  • The model of selling trivial items (e.g., lawn gnomes) via mail-based payments seemed preposterous to observers.
  • Founder Pierre Omidyar's pitch was perceived as chaotic, combining a hobbyist homepage about the Ebola virus with the nascent auction site.

Methodologies for Discovering Secrets

• Superior Tool Proficiency:
  • Dropbox: Drew Houston identified that while the market was saturated with cloud storage solutions, existing software was "janky" and slow (taking 10+ minutes to download).
  • The "secret" was realizing that true cloud storage required the reliability and speed of a local hard drive, a technical challenge solved by deep engineering expertise rather than business logic.
• Superior Problem Knowledge:
  • Kickstarter: CEO Perry Chen leveraged 10 years of experience within the arts community to identify a coordination failure in funding creative projects.
  • The model was dismissed initially as a bad idea because it relied on a patronage model (dating back to Renaissance Italy) that mainstream investors viewed as archaic or niche.
• Direct Life Experience:
  • SiteAdvisor: The founders identified a gap in security caused by social engineering (phishing, pop-ups) that incumbent firms like McAfee and Symantec ignored due to their focus on technical exploits.
  • Incumbents viewed the problem as a "user education" issue (white papers on being smarter), while the startup recognized it as a consumer experience problem.
  • The founders noted that convincing investors of obvious truths derived from experience is often frustrating because outsiders cannot see the problem through their established categorical filters.

Key Characteristics of "Good Ideas That Look Bad"

• Dismissal by Power as "Toys":
  • Telephone: Western Union dismissed the early telephone as a toy because it had limited range (approx. one mile) and failed to solve immediate railroad communication needs compared to the telegraph.
  • Skype: Early investors and incumbents dismissed VoIP as a toy due to low adoption of microphones, poor call quality, and skepticism about broadband infrastructure.
  • The Innovator's Dilemma: Technology improvements often outpace user demand initially, causing powerful incumbents to underestimate the trajectory of a "toy" technology.
• Unbundling of Functions:
  • Newspapers: The industry was dismantled piece-by-piece as digital entities unbundle its functions: distribution (internet), curatorial value (social media), and classified ads (Craigslist), and eventually individual reporter branding.
  • Higher Education (MOOCs): Platforms like Udacity and Coursera are unbundling the university experience, which consists of social interaction, credentialing, coursework, and recruitment.
  • The Pattern: Mainstream pundits dismiss each unbundled layer individually as a "stupid" replacement for the whole, failing to see the long-term trajectory of the total unbundling.
• Origins in Hobbies:
  • Technical innovators often "vote with their time" on weekends and nights rather than "allocating money," leading to innovations like the PC (Homebrew Computer Club), blogs, and the web.
  • GitHub: Began as an open-source tool for hobbyists but solved the same coordination needs required by large, semi-structured organizations (e.g., eBay).
  • Current hobbyist trends like Bitcoin, 3D printing, and drones are similarly dismissed as toys by mainstream business observers.
• Challenging Social Norms:
  • Flickr: Disrupted norms by defaulting photos to public viewing, whereas prior services (e.g., Shutterfly) operated on private repositories.
  • Facebook: Changes such as the Newsfeed algorithm update initially faced public outcry for challenging social sharing norms before becoming standard UI conventions.

Strategic Conclusions for Entrepreneurs

• Source of Ideas: The highest probability of finding "secrets" lies in direct experience (technical expertise, domain knowledge, or personal history) rather than abstract analysis.
• Failure of Abstract Trends: Ideas derived from analyst reports, trends, or analogies (e.g., "Uber for X") are often encapsulations of conventional wisdom and therefore lack the contrarian edge required for disruption.
• The Diff Formula: The most valuable startup ideas emerge when direct personal experience is subtracted from (diffed with) conventional wisdom.
• Prediction Validation: Systematic review of thousands of pitches and investments indicates that founder technical or problem-domain expertise is the strongest predictor of success, as it is the mechanism by which "secrets" are uncovered.