How to kill your bad ideas, w/Zynga's Mark Pincus

Episode Summary

Executive Summary: Reid Hoffman frames the episode around Mark Pincus’s core lesson: entrepreneurs must relentlessly pursue big opportunities while quickly killing bad ideas. Using Edison, Tribe, and Zynga, the episode shows how Pincus built a repeatable system for rapid experimentation, data standardization, and customer-centered product decisions, turning Zynga into a fast-learning machine.

Main Topics: Kill bad ideas fast, preserve winning instincts (Priority: 5/5): The central thesis is that entrepreneurs should be ruthless about abandoning ideas that don’t work while staying faithful to the underlying opportunity they sense. Mark Pincus’s early experiments and failures (Priority: 5/5): Pincus’s attempts at free PCs and the Tribe social network illustrate how even strong instincts can be attached to the wrong product concept. Zynga’s rapid testing culture (Priority: 5/5): Zynga used constant experimentation, quick user feedback, and frequent feature kills to identify hits like Farmville and Words with Friends. Data standardization as a scaling tool (Priority: 4/5): To scale decision-making, Zynga standardized metrics across teams so tests and results could be compared, shared, and acted on quickly. Bold beats and customer delight (Priority: 4/5): Small, low-risk feature changes could create outsized engagement when they delighted users, reinforcing the value of pixel-level product care. The limits of scale without focus (Priority: 4/5): Pincus argues that scaling should never distract from the actual customer experience; internal process matters only insofar as it improves delight.

Key Arguments: A strong entrepreneurial instinct is not enough; many ideas must be tested and killed quickly to find the right path. The Tribe failure proved that even when underlying instincts are correct, packaging multiple ideas into one product can doom mainstream adoption. Zynga’s advantage came from running hundreds of small tests, measuring user response immediately, and promoting only features that clearly resonated. Standardizing metrics across teams is essential because without common definitions, data becomes noise instead of decision-making fuel. Hiring analysts and data engineers early can be a strategic advantage when a company depends on constant experimentation. Customer experience should be optimized at the pixel level; users care about delight, not a company’s scaling mechanics. Small feature changes can become major growth drivers when they create visible emotional response and usage lifts.

Data Points: Ideas tested per day: hundreds - Zynga constantly tested ideas and quickly killed weak ones Free PC idea market response: nobody wanted a free PC - Mark Pincus’s early internet-access concept failed in street research Tribe launch year: 2003 - Pincus’s social network Tribe launched in the early social-network boom Social networks cited as contemporaries: MySpace, Tribe, LinkedIn, High Five, Friendster, Facebook - Reid notes 2003–2004 as a banner period for social networking Instinct vs. idea estimate: 95% / 25% - Pincus claims entrepreneurial instincts are right about 95% of the time while ideas may be right only 25% of the time Game development cycle comparison: 4 weeks vs. 2 years - Zynga could ship on the web in weeks while traditional game companies took up to two years Traditional game development spend: $15–20 million - Cost and time required before traditional companies knew if a game would work Feature click threshold: half of audience clicks - Pincus said a feature was likely worth pursuing if about half the audience clicked Farmville feature turnaround: a few hours - The moving cow feature was coded quickly and triggered major engagement Zynga data team size: 50 people vs. 10; 300 people vs. 20 or 50 - Pincus describes overinvesting in analytics compared with competitors Business growth claim for PEO use: twice as fast - Sponsor mention for DEEL citing the National Association of PEOs Capital One investment size: $40,000–$45,000 - Emily Warden invested upfront in a diamond collection with business card support

Pivotal Quotes: "I’ll try anything and I’ll kill anything and I’ll kill it quickly." — Mark Pincus: Summarizes his operating philosophy after the Tribe experience "If you don’t standardize data, you don’t have data. I mean, you just have a huge mess." — Andrea Jones-Rooy: Explains why common metrics are essential for meaningful analysis "I want our teams to care about the product experience at the pixel level." — Mark Pincus: Describes Zynga’s obsession with user experience and detail

Implications: Founders should build systems for fast learning: test constantly, standardize metrics, and kill weak ideas without hesitation. In competitive markets, speed plus disciplined data can turn small user insights into breakout products.