The Zebra Design Sprint

One flow. One week. Validated and shipped to your codebase.

This is a Design Sprint 4.0 — the fourth evolution of the methodology, rebuilt for an era where you ship code, not Figma. Every element in the process exists from feedback over 60+ client projects using this methodology.

The Lineage

Version	Who	Era	What Changed
1.0	Jake Knapp / Google Ventures	2010	The original — Map, Sketch, Decide, Prototype, Test in 5 days
2.0	AJ&Smart (Berlin)	~2016	Condensed to 4 days, remote-friendly, facilitation-first
3.0	Deep Work Studio (Charlie (me!) + Andrej)	2019	Remote first, with deliverables for early-stage startups
4.0	Zebra Design (Charlie)	2025	AI-era execution — one flow shipped in code not Figma and user-validated in ~5 days

Each generation got closer to the real thing. Sketches became prototypes. Prototypes became Figma. Figma became shipped code. Shipped code became validated with real users.

Who's it for

After iterating on the process with 60+ clients, the process is a perfect fit for pre-seed to seed stage startups needed the best in user experience and product design without hiring a dedicated designer.

Example One Sprint Structure

PRE-SPRINT: Prep & Onboarding
    │   Review learnings, onboard attendees, technical access
    ↓
WORKSHOP: Alignment / Feedback / Scope Lock
    │   Align stakeholders, involve domain experts,
    │   eliminate feedback loops before they start
    ↓
BUILD: Days 2-5
    │   Work directly in client codebase
    │   Ship working flow, scope lock mid-sprint
    ↓
DELIVERABLE: Working flow + user test ready
    │
    ════════════════════════════════════
         ASYNC GAP (Client's Timeline)
         Users complete tests over days/weeks
    ════════════════════════════════════
    │
    ↓
USER TESTING: ICP-weighted analysis
    │   Evidence over opinion
    ↓
UX FIXES: One-day iteration
    │   Fix what testing revealed
    │   Before next workshop or final presentation
    ↓
PRESENTATION: End of project
    │   Evidence-first results, deliverables, next steps
    ↓
VALIDATED OUTCOME

Clients can work in multiple sprints for multiple user-flows, features and user validation rounds.

The Process

Each step is a module — use the ones that fit where you are in the project. Click through for the detailed guide.

Guide	What It Covers
Project Setup	Review learnings, attendee selection, onboarding
Workshops	Alignment, scope lock workshop, feedback workshops — modular by sprint
Build & Prototype	Code before consensus, async communication, sprint protection rules
User Testing	Test protocol, ICP-weighted analysis, evidence over opinion
UX Fixes	Focused iteration day — fix what testing revealed
Presentations	Evidence-first results, deliverables, closing the engagement

Principles

Code before consensus — Build working software, get feedback on reality
The structure says no — Scope freeze, time bounds, and process create boundaries
Data decides — User testing as endpoint, not subjective approval
Tested, fixed, improved — UX fixes turn findings into shipped improvements before presenting
Recovery is methodology — Sustained quality requires structural rest

Why Zebra Sprints Exist

Design and code are being commoditised. Anyone can build an app now. The sustainable differentiator is knowing what to build, with UX working beautifully for humans and proving it works.

First engagement of the Design Sprint 4.0: 3 sprints, 55k lines of code, 173 components, 25+ screens. 9 user interviews across 2 rounds. Client: "I'm amazed. Unbelievable."

The sprint model delivers market value at a fraction of the cost because AI makes implementation fast — but knowing what to implement and validating it with real users is what saves months of wrong-direction engineering.