Lenny's PodcastJake Knapp & John Zeratsky: How a 2-day sprint sorts ideas
Note-and-vote silence and a Mad Libs founding hypothesis force a two-by-two map; teams lock customer differentiators before AI churns out a generic prototype.
CHAPTERS
- 0:00 – 5:54
Why startups fail early: misaligned customers and untested basics
Jake and John open with the most common early-stage breakdown they see: teams can’t clearly name their target customer, and even when they can, they don’t test their assumptions. This sets the stage for why a structured “Foundation Sprint” exists at all.
- •Founders often disagree on who the target customer really is
- •Two failure modes: unclear fundamentals vs. never validating them
- •The conversation will be highly tactical—focused on repeatable steps
- 5:54 – 9:29
From Design Sprint to Foundation Sprint: the Google Meet origin story
Jake recounts how clearing a week to build and test a prototype in Stockholm helped save a stalled project that became Google Meet. That experience evolved into the Design Sprint, later refined at Google Ventures with hundreds of teams.
- •Clearing calendars enabled deep focus and decisive progress
- •Prototype-first alignment beat PRDs and perfectionism
- •Design Sprint structure: map, sketch, decide, prototype, test
- •Scaling the method across startups and large companies at GV
- 9:29 – 11:42
Why Foundation Sprint exists: what was missing for pre-seed teams
John explains that early-stage, pre-product startups often lack the foundational clarity required to get full value from a Design Sprint. The Foundation Sprint was created to answer critical pre-work questions about customer, problem, competition, and differentiation.
- •Later-stage teams already know positioning and differentiation
- •Pre-seed teams often can’t crisply define the problem or customer
- •Foundation Sprint built at Character Capital (2021–2022)
- •Designed specifically for the beginning of high-risk projects
- 11:42 – 14:48
Foundation Sprint overview: clear 10 hours to align on the founding hypothesis
They outline the format: gather the core leadership team, clear calendars, and follow a scripted sequence of activities over ~10 hours (usually two days). The output is a concrete hypothesis, which then gets tested via Design Sprints over subsequent weeks.
- •Core team participates: product, engineering, design, marketing leadership
- •10-hour timebox (often two 4–6 hour blocks)
- •Creates clarity on basics, differentiation, and approach
- •Followed by 2–3+ weeks of Design Sprints to run experiments
- 14:48 – 17:15
Phase 1—Basics: customer, problem, competition, and advantages (with note-and-vote)
Jake breaks down the first phase: answer the “embarrassingly simple” questions that teams routinely gloss over. The key mechanic is ‘work alone together’—silent writing, voting, and a designated decider to force alignment fast.
- •Basics questions: most important customer, core problem, competitors and alternatives
- •Include ‘advantages’: insight, motivation, special capabilities
- •Method: silent ideation → vote → decider locks the answer
- •Speed matters: don’t over-invest here; save energy for differentiation
- 17:15 – 28:55
Case study—Latchit: capturing the basics for artisans selling online
They walk through Latchit (ex-Substack engineers) using the Basics template to clarify who they’re for and what they’re solving. The example highlights how even obvious-seeming answers become sharper—and more surprising—when forced into specifics.
- •Target customer: artisans who want to sell online but find tech/marketing hard
- •Problem: sales growth outside local community
- •Competition includes Shopify/Etsy plus offline alternatives (e.g., art fairs)
- •Advantage: founders previously built Substack’s network growth feature
- •Engineers’ instinct to build is tempered by upfront clarity work
- 28:55 – 40:17
Phase 2—Differentiation: building a customer-centered promise that stands out
Differentiation is framed as the heart of the Foundation Sprint: customers ignore new products unless the promise is distinct and compelling. The team generates and scores differentiators, then synthesizes them into a clear 2x2 positioning chart.
- •Differentiation must be from the customer’s perspective (not investor tech framing)
- •Start with classic differentiators (fast/slow, easy/hard, focused/broad, etc.)
- •Write custom differentiators unique to the product’s worldview
- •Score honestly against competitors to avoid wishful thinking
- •Create a 2x2 chart that pushes competitors into “Loserville” quadrants
- 40:17 – 43:38
Price as a differentiator (and why it’s usually weak—except in some AI cases)
Lenny presses on pricing strategy as differentiation. John argues price is rarely durable because incumbents can outspend and subsidize; AI can sometimes unlock real cost structure advantages, but it often must be dramatically cheaper to offset quality tradeoffs.
- •Competing on price is difficult vs. large incumbents and loss leaders
- •Price advantages can be real when AI replaces previously manual work
- •Example: Bindwell uses AI to design pesticides, enabling lower costs
- •Rule of thumb discussed: AI offerings may need ~10x cheaper to win early
- 43:38 – 49:45
Custom differentiators in practice: Mellow and Latchit’s final picks
They show how teams generate many candidate differentiators, then narrow to one or two that are both deliverable and valuable. Mellow and Latchit illustrate the discipline of not trying to win every axis—only the ones that matter most.
- •Custom differentiators can be subtle wording shifts that change perception
- •Mellow context: focused, high-quality AI agents vs. overpromising tools
- •Mellow’s chosen differentiators: “mobile first” and “works out of the box”
- •Latchit’s chosen differentiators: “helps you grow” and “cooperative”
- •Honest scoring creates clearer positioning and marketing claims
- 49:45 – 52:02
The mini-manifesto: turning differentiation into decision-making principles
To make differentiation operational, teams translate it into project principles—short rules that guide tradeoffs during building. The output is a one-page ‘mini-manifesto’ combining the positioning map and principles that keep execution aligned.
- •Principles help arbitrate product decisions quickly and consistently
- •Example principles for Latchit: ‘Help sellers help each other’ and ‘Do the thing that makes sellers more money’
- •Inspired by Google’s ‘10 things we know to be true’ (e.g., ‘Fast is better than slow’)
- •Mini-manifesto acts as a practical North Star, not a pitch-deck gimmick
- 52:02 – 1:17:30
Phase 3—Choosing the approach: “Magic Lenses” to pick a path (and a backup)
With basics and differentiation set, the team evaluates multiple implementation approaches using “Magic Lenses”—different viewpoints like customer value, pragmatism/speed, growth, financial upside, and differentiation. This produces a clear first-choice approach plus a backup plan to reduce fear and enable faster pivots.
- •Approaches are alternate paths to the same destination (not new company ideas)
- •Lenses include: customer expert, pragmatic (cheap/fast), growth, money, differentiation
- •Teams add custom lenses like founder excitement/conviction (“F yeah” vs “Nah”)
- •Visual plotting reveals either a clear winner or unavoidable tradeoffs
- •Outcome: first choice + backup plan
- 1:17:30 – 1:19:34
Founding hypothesis: the one-sentence strategy you can finally test
All three phases culminate in a ‘Mad Libs’ hypothesis statement that makes implicit strategy explicit. With the hypothesis written down, teams can interrogate it, test each variable, and prevent hidden misalignment from derailing execution.
- •Template: customer + problem + approach + competitors + differentiators
- •Latchit example: artisans → online sales growth → social sales app; win vs. Shopify/Etsy due to cooperative + easy to use
- •Hypothesis framing reinforces that this is a testable belief, not a guarantee
- •Making assumptions explicit enables clearer experiments and team alignment
- 1:19:34 – 1:28:51
Design Sprints after the Foundation: prototypes, scorecards, and rapid iteration to ‘click’
They explain how Design Sprints operationalize the hypothesis: identify biggest risks, prototype key moments, and test with real customers. A new addition is the ‘scorecard,’ which breaks down hypothesis components and tracks what’s working across interviews and sprint cycles.
- •Start each sprint by identifying biggest risks to the hypothesis
- •Prototype the ‘key moment’ (e.g., landing pages) and test head-to-head variants
- •Use a scorecard to evaluate customer fit, problem validity, approach, differentiation, and competitive choice
- •Latchit progression: early sprints show lots of red, later sprint turns fully green
- •Founders report compressing 3–4 months of learning into 3–4 weeks
- 1:28:51 – 1:37:10
AI and prototyping: speed helps, but generic outputs punish shallow thinking
They explore how AI changes sprint execution: teams can create more realistic prototypes faster, including videos and partial implementations. The risk is that AI-generated work tends to be generic unless the team does the hard differentiation and messaging thinking first.
- •Primary AI use: faster, more realistic prototypes (“prototyping team on standby”)
- •Key warning: don’t outsource the thinking—especially messaging and differentiation
- •Example: Axion Orbital uses marketing pages + demo video to simulate complex product behavior
- •Sketching/detail planning acts like ‘prompt engineering’ for better vibe-coded outcomes
- •Going fast too early can slow you down by locking you into generic directions
- 1:37:10 – 1:41:33
Motivation, resources, and how to run it yourself (template + Character Labs)
They close with the broader value: sprints bring teams closer to customers and to each other through focused, structured work. They share where to get the Miro template and how to learn more about Character Labs and their work.
- •Structured, calendar-cleared work reduces social dynamics and context switching
- •Weekly customer contact builds empathy and momentum
- •Miro Foundation Sprint template is available via character.vc
- •Jake and John can be found on LinkedIn; founders can apply to Character Labs
- •Final recap: beginnings matter—clarity first, then speed
