YC Root Access

Lecture 8 - How to Get Started, Doing Things that Don't Scale, Press

Lecture Transcript: http://tech.genius.com/William-sydney-walker-lecture-8-doing-things-that-dont-scale-pr-and-how-to-get-started-annotated Lecture 8 features 3 speakers: Stanley Tang, Founder of Doordash, covers How to Get Started. Walker Williams, Founder of Teespring, covers Doing things that Don't Scale. Justin Kan, Founder of TwitchTV and Partner at Y Combinator, covers Press. See the slides and readings at startupclass.samaltman.com/courses/lec09/ Discuss this lecture: https://startupclass.co/courses/how-to-start-a-startup/lectures/64037 This video is under Creative Commons license: http://creativecommons.org/licenses/by-nc-nd/2.5/

STStanley TangguestWWWalker WilliamsguestJKJustin Kanguest

Oct 16, 2014·52m·Watch on YouTube ↗

📖 CHAPTERS

1. 1

   Stanley Tang’s Stanford-to-Series A snapshot: the “ridiculous” early startup life

   Stanley Tang (DoorDash cofounder) opens by contrasting student life with suddenly running a venture-backed startup. He uses a vivid moment—walking back to his Stanford dorm after signing a $15M Series A—to frame how fast the journey can move.

2. 2

   Discovery in a macaron shop: identifying delivery as a small-business pain point

   DoorDash’s initial spark came from interviewing a macaron shop owner about problems facing small businesses. Delivery demand showed up as a repeat, urgent constraint—orders existed, but merchants lacked a way to fulfill them.

3. 3

   Customer validation at scale: 150–200 interviews and the “why hasn’t this been solved?” question

   The team expanded interviews to hundreds of local merchants and repeatedly heard the same delivery problem. This raised a critical founder question: if it’s so obvious, what’s missing—especially around true consumer demand?

4. 4

   The scrappy experiment: launching PaloAltoDelivery.com in an afternoon

   Instead of building infrastructure, they created a barebones landing page with PDF menus and a phone number that routed to their personal cell phones. The goal was simple: see if real orders would happen with minimal effort.

5. 5

   First order reality check: delivering pad thai themselves

   A real phone order arrived almost immediately, forcing the team to operationalize on the spot. They personally picked up and delivered the food, learning firsthand what a delivery workflow actually entails.

6. 6

   Early traction despite a terrible UX: signals of strong demand

   Orders ramped from one to multiple per day even though the user experience was clunky and unpolished. The willingness of customers to tolerate friction became a strong product-demand signal.

7. 7

   Doing things that don’t scale (DoorDash): founders as drivers, dispatch, and support

   Stanley emphasizes the YC mantra of doing things that don’t scale: founders handled delivery, support, and marketing manually. These hacks weren’t just stopgaps—they built deep intuition for what to automate later.

8. 8

   Manual work as a learning engine: becoming experts and shaping automation

   Operating manually taught the team the true constraints of logistics, customer expectations, and dispatch decisions. They also built loyalty through personalized outreach, turning early users into advocates.

9. 9

   Stanley’s three takeaways + Q&A: distribution, mobile, competition, and long-term vision

   Stanley closes with a tight framework—test hypotheses, launch fast, and embrace non-scalable tactics—then answers tactical questions. He highlights mobile as the enabling shift and describes DoorDash’s broader merchant-delivery ambition.

10. 10

    Walker Williams (Teespring): why non-scalable tactics are a startup’s unfair advantage

    Walker defines “things that don’t scale” as fundamentally unsustainable early growth strategies that won’t reach the millionth user—but can bootstrap the first thousand. He structures the talk around acquiring first users, turning them into champions, and finding product-market fit.

11. 11

    Getting first users: there’s no silver bullet (and early sales will look terrible)

    Walker describes Teespring’s early days as painfully manual—free design help, revisions, and hands-on launches to sell small batches. He warns founders not to optimize for ROI early and to resist using “free” as a crutch that distorts real demand.

12. 12

    Turning users into champions: daily user contact, service recovery, and social listening

    Walker argues that memorable support and proactive communication create champions who drive growth. He recommends founders stay close to user feedback through hands-on support, churn outreach, and constant monitoring of social channels.

13. 13

    Finding product-market fit fast: optimize for speed over clean architecture

    Walker encourages shipping fast, even if it creates technical debt, because early iterations will change the product anyway. He shares examples of shortcut engineering and operational “pain” (crashes, paging) as a tradeoff to reach fit sooner.

14. 14

    Q&A with Walker: competing in a crowded market and evolving customer segments

    Walker explains Teespring’s origin in a personal pain point and notes that “silly” ideas can work if adoption shows strong pull. He also clarifies that Teespring grew beyond nonprofits into entrepreneurs and influencers building brands.

15. 15

    Justin Kan on press: set goals, pick the right story, and run PR like a funnel

    Justin demystifies press as non-magical and non-meritocratic: it’s driven by incentives, relationships, and clear objectives. He explains how to align coverage targets with business goals and how to package “newsworthy” stories.

16. 16

    Tactical PR playbook + PR firms: warm intros, lead time, structured pitches, and relationship-building

    Justin outlines a step-by-step process: get warm introductions, meet reporters with lead time, and control the narrative with a prepared outline. He cautions that PR firms are expensive and can’t manufacture what’s inherently interesting, and reminds founders that press is a short-term bootstrap channel.