What is The Crossley Method?

The Crossley Method is a seven-stage framework for taking a product from first idea to first revenue: Discover, Validate, Prototype, Build MVP, Launch, First Revenue and Scale. Each stage has defined deliverables and an explicit kill-or-continue decision, so the cheapest possible evidence is gathered before the most expensive work begins.

How long does the whole method take?

Six to eight weeks from a standing start to first revenue for a focused software product. Two weeks of discovery and validation, one week of prototyping, two to three weeks of build, one week to launch, and the revenue stage running alongside launch. Scale is ongoing.

Can I skip the validation stages?

You can, and it is the most expensive shortcut available. Two weeks of validation costs a few thousand pounds; discovering the same information after a build costs £12,000 to £25,000 and several months. Skip validation only when you already have behavioural evidence — paying customers or signed commitments.

Does the method work for AI products?

Yes, with one addition: the evaluation set is written during Validate rather than after Build. If you cannot write fifty real examples with the behaviour you would consider correct, the AI feature is not yet specified well enough to build.

What makes this different from Lean Startup?

It shares the same lineage but is more prescriptive. Each stage names the deliverables, the timebox and the specific evidence threshold that permits moving on, which is the part founders most often improvise and most often get wrong.

Who is the method for?

Non-technical and technical founders at pre-seed and seed stage building software products, particularly AI SaaS. It also works as an operating framework for a product team that keeps shipping without moving the numbers.

Andrew Crossley

THE FRAMEWORK

The Crossley Method

Seven stages from first idea to first revenue: Discover, Validate, Prototype, Build MVP, Launch, First Revenue, Scale. Every stage has defined deliverables, a timebox and an explicit kill-or-continue decision — so the cheapest evidence gets gathered before the most expensive work starts.

STAGE 1

Discover

Week 1

Discovery is not idea generation. It is finding a group of people who already have an expensive problem and understanding what they do about it today. Fifteen conversations inside one tightly defined segment, asking about past behaviour rather than future intentions.

DELIVERABLES

—Defined segment, narrow enough to name individuals
—Problem statement in the customer's own language
—The current workaround and what it costs them

Kill or continue: Continue only if the workaround costs real money, time or risk. A mild annoyance never becomes a business.

Work with me on this stage →

STAGE 2

Validate

Week 2

Validation means designing a test where a yes costs the customer something. Pre-orders, paid pilots, signed letters of intent, or sustained use of a manual concierge version. Thresholds are agreed before the test runs.

DELIVERABLES

—A live demand test with real pricing
—Behavioural evidence, or a documented no
—A written kill-or-continue recommendation

Kill or continue: Three independent behavioural commitments from one segment is the bar. Below it, you are funding a hypothesis.

Work with me on this stage →

STAGE 3

Prototype

Week 3

Make the core journey clickable before it is buildable. Five real users walk through it. It always changes, and changing a prototype costs hours where changing a build costs weeks.

DELIVERABLES

—Clickable prototype of one journey
—Five user walkthroughs with notes
—A scoped build brief and an explicit not-building list

Kill or continue: Continue when five users complete the journey without narration from you.

Work with me on this stage →

STAGE 4

Build MVP

Weeks 3-4

One journey, deployed from week one, built daily with AI tooling. Auth, data model, the core loop and nothing else. Analytics and, for AI features, an evaluation set go in alongside the code rather than afterwards.

DELIVERABLES

—Live product on a real URL
—Activation funnel instrumented
—Evaluation set for any AI behaviour

Kill or continue: Continue when a non-technical stranger completes the journey unaided.

Work with me on this stage →

STAGE 5

Launch

Week 5

Launch is a surface, not an event: landing page, pricing, onboarding, and a working payment or waitlist path. Most MVPs that fail at this stage fail because there was nowhere for interest to convert.

DELIVERABLES

—Landing page with real pricing
—Onboarding sequence
—Payment or waitlist live

Kill or continue: Continue when strangers, not friends, are signing up.

Work with me on this stage →

STAGE 6

First Revenue

Week 6

Revenue is the only honest validation. The goal of this stage is the first paid customer who is not doing you a favour, and a clear read on what they bought and why.

DELIVERABLES

—First paying customers
—Confirmed pricing and packaging
—Documented reason customers bought

Kill or continue: Continue to scale only when the reason people pay is repeatable and you can name it.

Work with me on this stage →

STAGE 7

Scale

Ongoing

Scale is an operating model problem, not a feature problem. Decision rights, planning cadence, a defined metric set, protected discovery time and the product team you need to repeat the result without you in every room.

DELIVERABLES

—One-page product strategy
—Twelve-month roadmap across three horizons
—Operating cadence and hiring plan

Kill or continue: Ongoing. The measure is whether good decisions happen when the founder is not present.

Work with me on this stage →

THE METHOD IN PRACTICE

Wocal, from one venue to 300+

Wocal started as a manual service for a single venue. That concierge period was the Discover and Validate stages in everything but name: it revealed the actual job the venue needed doing, which was narrower and more specific than the hospitality platform originally imagined.

The build stage shipped one journey rather than a suite. Launch came with pricing attached from the start, which forced the uncomfortable question — will anyone pay — into week six instead of month nine. From 2020 to 2025 that discipline took the business to 300+ venues and a £2.7M pre-money valuation.

Co-Ride, founded in November 2025, is running the same sequence from the start, and CallFlow AI, Journly and Vynl were each scoped down to a single journey before a line of production code was written.

FREQUENTLY ASKED

Questions about the method

What is The Crossley Method?: The Crossley Method is a seven-stage framework for taking a product from first idea to first revenue: Discover, Validate, Prototype, Build MVP, Launch, First Revenue and Scale. Each stage has defined deliverables and an explicit kill-or-continue decision, so the cheapest possible evidence is gathered before the most expensive work begins.
How long does the whole method take?: Six to eight weeks from a standing start to first revenue for a focused software product. Two weeks of discovery and validation, one week of prototyping, two to three weeks of build, one week to launch, and the revenue stage running alongside launch. Scale is ongoing.
Can I skip the validation stages?: You can, and it is the most expensive shortcut available. Two weeks of validation costs a few thousand pounds; discovering the same information after a build costs £12,000 to £25,000 and several months. Skip validation only when you already have behavioural evidence — paying customers or signed commitments.
Does the method work for AI products?: Yes, with one addition: the evaluation set is written during Validate rather than after Build. If you cannot write fifty real examples with the behaviour you would consider correct, the AI feature is not yet specified well enough to build.
What makes this different from Lean Startup?: It shares the same lineage but is more prescriptive. Each stage names the deliverables, the timebox and the specific evidence threshold that permits moving on, which is the part founders most often improvise and most often get wrong.
Who is the method for?: Non-technical and technical founders at pre-seed and seed stage building software products, particularly AI SaaS. It also works as an operating framework for a product team that keeps shipping without moving the numbers.

RUN THE METHOD WITH ME

Where are you stuck?

Tell me the stage and what is blocking it. I reply within 48 hours.

EXPLORE NEXT

Continue the journey

Every stage from first idea to first revenue, mapped end to end.

The Crossley Method

Discover

Validate

Prototype

Build MVP

Launch

First Revenue

Scale

Wocal, from one venue to 300+

Questions about the method

Where are you stuck?

Continue the journey

Product Validation

MVP Development

Idea to MVP

Product Strategy Consultant

Fractional Chief Product Officer

MVP Cost Calculator