By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.
My First MillionHow a $5B founder is using AI (3 tutorials)
Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.
CHAPTERS
- 0:00 – 1:19
Cold open: “$5B founder,” Zapier pronunciation, and the API-in-the-name origin story
The hosts open with banter about clickbaiting the episode title and what Zapier is worth. Wade Foster explains how to pronounce “Zapier,” plus two origin details: the missing “p” in the domain and the hidden “API” inside the name.
- •The hosts riff on the “$5 billion founder” framing and valuation curiosity
- •Wade Foster explains “Zapier makes you happier” and how it rhymes with “happier”
- •Zapier’s name came from a domain constraint (couldn’t get the second “p”)
- •Wade notes the name also embeds “API,” reflecting the product’s purpose
- 1:19 – 2:13
What Zapier does and how the business got valued in the billions
Shaan explains Zapier’s core utility as “universal plugs” that move data between otherwise disconnected apps. Wade clarifies they were largely bootstrapped, raising only a small early round, and he confirms the $5B figure came from a secondary sale in 2021 with revenue up since then.
- •Shaan describes Zapier’s core value: connecting apps so information flows automatically
- •Wade describes their funding as “seed strap” and cites only about a million-plus raised
- •Wade attributes the ~$5B number to a 2021 secondary sale
- •Wade says revenue is up since that time
- 2:13 – 8:35
The AI-era sales pitch: connect your tools, then deploy deterministic automations
Wade pitches Zapier’s role in the AI workflow: chatbots are generally smart, but become business-specific when connected to your company tools. He argues the real power comes from deploying background automations/agents, with Zapier emphasizing more deterministic, reliable workflows that only invoke AI where needed.
- •Wade contrasts “AI as a chatbot” vs AI informed by your CRM, email, help desk, and chat tools
- •He frames the next step as deploying agents/automations that operate while you sleep
- •He argues many agents ‘guess’ and can be expensive due to token usage
- •He positions Zapier as writing workflow logic/code for reliability and cost-effectiveness
- •He notes natural-language building inside tools like Claude can be better than older no-code approaches
- 8:35 – 14:16
Robot executive staff, part 1: the morning brief (“Waking Up Like The President”)
Wade demos a simple starting agent: a daily morning briefing that pulls from calendars and other sources and sends a structured message (e.g., in Slack). Sam and Shaan compare notes on their own briefing routines and why having it run in the cloud matters.
- •Wade describes a 6:00 AM agent that compiles calendar and context into a Slack brief
- •Shaan calls it “Waking Up Like The President,” like a daily intelligence brief
- •Sam shares he has a version that prints and includes a pump-up message
- •Wade shows the Zapier workflow, emphasizing deterministic steps plus targeted AI usage
- •Wade explains the benefit vs laptop-dependent routines: Zapier runs it in the cloud
- 14:16 – 17:39
Robot executive staff, part 2: evening brief + ‘Scribe’ that logs your day and drafts follow-ups
Wade explains why an end-of-day (or post-meeting) workflow can be more valuable than a morning brief. His system loops through meeting notes, unfinished to-dos, and outstanding emails, prompts reflection on how the day went, and proactively drafts action-item follow-ups—cutting his end-of-day admin time dramatically.
- •Wade describes pulling meeting notes (he mentions using Granola), to-dos, and inbox items
- •The agent asks for a reflection on the day and logs what made it good/bad over time
- •It proposes and drafts follow-up actions (e.g., intros requested in meetings)
- •Wade says it reduced his end-of-day catch-up from about two hours to about 15 minutes
- •He uses summarized patterns (good-day vs bad-day traits) to help his assistant shape his schedule
- 17:39 – 19:27
AI that argues with you: system prompts to challenge assumptions, not flatter you
Wade shows how he configures his AI to be direct, disagree, and poke holes in his thinking. He demonstrates with a provocative example (killing a free tier) where the AI pushes back, asks for data, and suggests smaller experiments.
- •Wade describes maintaining a ‘how to work with me’ prompt file (e.g., agents.md)
- •He wants the AI to tell him truths coworkers may avoid and to challenge assumptions
- •He explains AIs are goal-seeking and need explicit guidance to avoid default placation
- •In his example, the AI responds by pausing the decision and requesting evidence and clarity
- •The hosts react positively to the pushback format as a leadership tool
- 19:27 – 27:41
The ‘War Council’ skill: seven sub-agents with personas + why Wade uses coding agents as chat partners
Wade introduces a “War Council” approach that spins up multiple sub-agents with assigned personas, including standing roles like a wartime COO and ruthless CFO, plus dynamically generated roles based on the prompt. He also explains why he often uses tools like Cursor as his default AI interface and discusses voice-to-text and the emerging ‘personality’ dimension in software.
- •Wade describes seven sub-agents/personas, with some standing members and others generated per prompt
- •He explains he defined personas via traits and then edited what the AI generated
- •He says he uses Cursor largely because it can take action (write code/do tasks), even though he doesn’t write much code
- •Wade mentions using voice-to-text via an app called Monologue
- •Conversation shifts to how software and models develop recognizable personalities and user preferences
- 27:41 – 31:33
CEO CRM + “who’s the boss?”: AI-driven customer outreach suggestions
Wade demos a CEO-focused CRM view showing account health signals and suggested outreach. He describes waking up to a weekly list of customers to contact, with draft emails prepared, and reflects on how AI can start directing the human’s priorities—so long as you keep your judgment engaged.
- •Wade shows an account view with usage changes, renewal timing, key contacts, and suggested outreach
- •He describes a weekly routine: a Sunday briefing with 10 customers to contact and drafted emails ready
- •He says this helps compensate for not being naturally inclined to keep in touch broadly
- •Wade frames it as AI becoming a ‘boss’ in certain contexts, assigning priorities
- •He cautions you still need to keep your brain on to avoid sounding dumb
- 31:33 – 43:34
AI in the loop for high-stakes decisions—especially hiring approvals
Wade explains that AI now has a ‘seat at the table’ for major decisions, and gives a detailed example from hiring. He contrasts pre-AI friction—where his intuition was hard to articulate and could feel like CEO override—with post-AI clarity, where the AI’s articulation improves the panel’s debate and decision quality.
- •Wade says AI weighs in on high-stakes decisions and is consulted in exec/board contexts
- •He describes Zapier’s bar-raiser-style offer approval process and his role in auditing it
- •Pre-AI, he could ‘smell’ issues but struggled to express why, creating tension with leaders
- •Post-AI, War Council outputs more articulate critique and comparisons, prompting better scrutiny
- •He observes people find it easier to debate the AI’s points than to argue directly with the CEO
- 43:34 – 48:15
Can small teams run huge businesses? Limits of agent fleets and the new management skill
Sam pushes the idea that large headcount may be replaced by agent fleets, and Wade answers in “unhinged mode.” Wade agrees managing agents will become a key skill, but argues there are cognitive limits and some roles—like sales—still benefit from humans focusing on relationships rather than only managing agents.
- •Sam asks whether one person could monitor many agents and radically shrink headcount
- •Wade argues there’s still a practical ‘span of control’ limit—even for managing agents
- •They discuss whether very large revenue outcomes could be achieved with very small teams in some business models
- •Shaan proposes compensation/value will increasingly track ability to manage effective agent fleets
- •Wade agrees on direction while noting some functions still prioritize human connection (e.g., sales)
- 48:15 – 57:02
Wealth, motivation, and defining winning: under-40 billionaire talk and life priorities
The conversation turns to Wade’s age, the ‘billionaire before 40’ framing, and how he thinks about money. Wade emphasizes family as his top priority and says beyond a certain threshold, more wealth doesn’t change the equation much; he shares a role-model story about his grandfather and the hosts discuss avoiding “1-dimensional winner” definitions.
- •Wade says the billionaire framing feels ‘fictional’ because much is paper value
- •He prioritizes family and says after a certain level, more money doesn’t change much
- •He recalls earlier goals like wanting to make $100k/year and notes he lacks expensive vices
- •Wade shares a detailed story about his grandfather’s life and community impact as a model of priorities
- •Shaan discusses being ‘as rich as what you don’t need’ and warns against copying one-dimensional winners
- 57:02 – 1:06:55
Unusual founder choices + what they’re reading and listening to
Wade gives examples of decisions that may look odd in a typical Silicon Valley playbook, including remote work from the start and raising minimal capital while focusing on customers and the work itself. The episode closes with media recommendations from all three, spanning podcasts and books.
- •Wade cites early remote work, minimal fundraising, and long-haul customer focus as distinctive choices
- •He argues the numbers follow when you focus on doing the work well
- •Wade recommends the Acquired podcast, naming episodes/series he likes (e.g., NFL; major retailers)
- •Wade recommends the book Make Something Wonderful (Steve Jobs in his own words)
- •Shaan recommends The Score; Sam recommends Mark Hoppus’s memoir Fahrenheit-182
