E38: My Worst Hires Almost Destroyed Us (Here's What I Learned About Building Teams in 2025)

My Worst Hires Almost Destroyed Us (Here's What I Learned)After scaling from 15 people down to 4, then back to 5, I've never shared my actual hiring criteria. Today I'm telling you about the engineer who took unlimited PTO during his wedding while our engineering delivery collapsed, the sales guy who learned to sell to us, and the contractor who treated us like an ATM.These aren't hypotheticals - these are real people who worked with us. Here's what I learned about the two types of bad hires every founder will encounter, and my framework for hiring in the AI-first world.Key Topics CoveredThe Best Hire Story (3:15)Meeting Rashdeep in CS 136 class at WaterlooFrom desk mates to co-founders in a basement conversationWhy genuine friendship created the foundation for an amicable co-founder separationClear division of responsibilities: CTO vs CEO rolesType 1 Bad Hire: The "I" Person (8:30)The engineer who used unlimited PTO as cover for continuous wedding leaveFamily emergency excuses that coincidentally aligned with honeymoon periodHow engineering delivery collapsed when he became unreachableWhy unlimited PTO policies are dangerous for startupsType 2 Bad Hire: Money-Only North Star (15:45)The San Diego sales guy who figured out George was "easy going"How he used high-pressure sales tactics on his own companyContinuous promotion demands using leverage and manipulationWhy George doesn't hire salespeople anymoreThe Great Simplification (22:15)From 5 engineers down to 1 in early 2023How AI covered the engineering capacity gapThe conversation that saved the remaining engineerWhy loyalty and culture fit trumped technical specializationThe Two Types of Bad Hires FrameworkType 1: The "I" PersonCharacteristics:Everything revolves around themDon't think of themselves as employees, but as entities profiting off youDo bare minimum to not get firedNo mission alignment - job hop without belief in visionMake decisions that benefit them at team's expenseRed Flags:Recommending friends for hiring queue for personal benefitSuggesting changes that hurt other employees to benefit themselvesTreating company like an ATM rather than a missionType 2: Money-Only North StarCharacteristics:Compensation is sole motivatorWill use leverage against you once they realize you need themTurn learned skills inward (sales tactics on own company)Manipulative promotion and raise requestsNo cultural alignment with team valuesRed Flags:High-pressure tactics during internal conversationsLeveraging "other offers" for continuous promotionsMaking unreasonable demands shortly after getting what they wantedFocus on individual compensation over team successGeorge's Modern Hiring FrameworkThe Fundamental ShiftOld Way: Hiring for roles (DevOps, Frontend, Backend specialists) New Way: Hiring for functions (engineering capacity, strategic thinking)Priority Split:50% Culture fit50% Capabilities (including AI leverage)Interview Process ChangesPre-AI Era:24-48 hour take-home coding exercisesTechnical skill assessmentsRole-specific questionsPost-AI Era:Open-ended problem-solving questions"No AI" policy for technical discussionsFocus on thinking process over coding abilityDebugging scenarios to understand decision-makingKey Interview Questions"What's your thinking process for this problem?""What's your relationship with AI? How do you use it?""Can you give us a concrete example?""How would you debug this real-world scenario?""Walk us through your decision-making process"The Brutal Lessons LearnedWhy George Fired the Sales Guy"For my career, I'd never feel good about letting someone go. But for this person, I was actually pretty happy to let them go. I felt a huge sense of relief."The sales person had turned his own skills against the company, using high-pressure tactics learned from car dealerships to manipulate continuous promotions.The Single Point of Failure ProblemHaving one sales person created leverage they exploited. George learned that essential roles need redundancy or internal capability.Why Unlimited PTO Failed"Everyone should have a specific number of days each year that can take off. And they need to give at least 24 hours notice."Current policy: 14-15 days PTO for all employees, increasing with tenure.Cultural Screening FrameworkThe Core Principle"Every single bad decision I made could be saved with rigorous cultural screenings."The Screening ProcessAsk open-ended questions about problem-solvingHave them solve problems in front of you, with youAsk the same question in different ways to check consistencyLook for alignment between stated values and behaviorRed Flag Detection"It's very easy to conceal temporarily, but impossible to conceal forever."Use the police interrogation method: ask similar questions multiple times in different ways to find inconsistencies.The AI-Era Hiring RealityWhat's ChangingAmazon: 30,000 layoffs (mostly corporate/middle management)Meta: 10,000+ job cutsAI replacing KPI tracking and performance management layersGeorge's Prediction"We're moving to a world where jobs as we know don't exist anymore."The Shift:From "it's not my job" to task-based workFrom departmental roles to function-based outcomesFrom monthly salaries to per-task payment (DoorDash economy)Future Team Structure"SimpleDirect might be George + AI + two really excellent humans who can think strategically and work with AI as leverage."When to Actually HireGeorge's Strict CriteriaOnly hire when:Revenue skyrockets and you literally cannot serve customersExisting team + AI physically cannot handle more (emphasis on "impossible")You're turning away money, not just feeling overwhelmedWhat Doesn't Justify HiringFeeling overwhelmedThinking you "should" have someone in that roleTeam seems busyRevenue growth without capacity constraintsTeam Management PhilosophyGlobal Team ApproachEmployees/contractors worldwidePhysical...